1 | //===-- AMDGPULowerBufferFatPointers.cpp ---------------------------=// |
2 | // |
3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
4 | // See https://llvm.org/LICENSE.txt for license information. |
5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
6 | // |
7 | //===----------------------------------------------------------------------===// |
8 | // |
9 | // This pass lowers operations on buffer fat pointers (addrspace 7) to |
10 | // operations on buffer resources (addrspace 8) and is needed for correct |
11 | // codegen. |
12 | // |
13 | // # Background |
14 | // |
15 | // Address space 7 (the buffer fat pointer) is a 160-bit pointer that consists |
16 | // of a 128-bit buffer descriptor and a 32-bit offset into that descriptor. |
17 | // The buffer resource part needs to be it needs to be a "raw" buffer resource |
18 | // (it must have a stride of 0 and bounds checks must be in raw buffer mode |
19 | // or disabled). |
20 | // |
21 | // When these requirements are met, a buffer resource can be treated as a |
22 | // typical (though quite wide) pointer that follows typical LLVM pointer |
23 | // semantics. This allows the frontend to reason about such buffers (which are |
24 | // often encountered in the context of SPIR-V kernels). |
25 | // |
26 | // However, because of their non-power-of-2 size, these fat pointers cannot be |
27 | // present during translation to MIR (though this restriction may be lifted |
28 | // during the transition to GlobalISel). Therefore, this pass is needed in order |
29 | // to correctly implement these fat pointers. |
30 | // |
31 | // The resource intrinsics take the resource part (the address space 8 pointer) |
32 | // and the offset part (the 32-bit integer) as separate arguments. In addition, |
33 | // many users of these buffers manipulate the offset while leaving the resource |
34 | // part alone. For these reasons, we want to typically separate the resource |
35 | // and offset parts into separate variables, but combine them together when |
36 | // encountering cases where this is required, such as by inserting these values |
37 | // into aggretates or moving them to memory. |
38 | // |
39 | // Therefore, at a high level, `ptr addrspace(7) %x` becomes `ptr addrspace(8) |
40 | // %x.rsrc` and `i32 %x.off`, which will be combined into `{ptr addrspace(8), |
41 | // i32} %x = {%x.rsrc, %x.off}` if needed. Similarly, `vector<Nxp7>` becomes |
42 | // `{vector<Nxp8>, vector<Nxi32 >}` and its component parts. |
43 | // |
44 | // # Implementation |
45 | // |
46 | // This pass proceeds in three main phases: |
47 | // |
48 | // ## Rewriting loads and stores of p7 |
49 | // |
50 | // The first phase is to rewrite away all loads and stors of `ptr addrspace(7)`, |
51 | // including aggregates containing such pointers, to ones that use `i160`. This |
52 | // is handled by `StoreFatPtrsAsIntsVisitor` , which visits loads, stores, and |
53 | // allocas and, if the loaded or stored type contains `ptr addrspace(7)`, |
54 | // rewrites that type to one where the p7s are replaced by i160s, copying other |
55 | // parts of aggregates as needed. In the case of a store, each pointer is |
56 | // `ptrtoint`d to i160 before storing, and load integers are `inttoptr`d back. |
57 | // This same transformation is applied to vectors of pointers. |
58 | // |
59 | // Such a transformation allows the later phases of the pass to not need |
60 | // to handle buffer fat pointers moving to and from memory, where we load |
61 | // have to handle the incompatibility between a `{Nxp8, Nxi32}` representation |
62 | // and `Nxi60` directly. Instead, that transposing action (where the vectors |
63 | // of resources and vectors of offsets are concatentated before being stored to |
64 | // memory) are handled through implementing `inttoptr` and `ptrtoint` only. |
65 | // |
66 | // Atomics operations on `ptr addrspace(7)` values are not suppported, as the |
67 | // hardware does not include a 160-bit atomic. |
68 | // |
69 | // ## Type remapping |
70 | // |
71 | // We use a `ValueMapper` to mangle uses of [vectors of] buffer fat pointers |
72 | // to the corresponding struct type, which has a resource part and an offset |
73 | // part. |
74 | // |
75 | // This uses a `BufferFatPtrToStructTypeMap` and a `FatPtrConstMaterializer` |
76 | // to, usually by way of `setType`ing values. Constants are handled here |
77 | // because there isn't a good way to fix them up later. |
78 | // |
79 | // This has the downside of leaving the IR in an invalid state (for example, |
80 | // the instruction `getelementptr {ptr addrspace(8), i32} %p, ...` will exist), |
81 | // but all such invalid states will be resolved by the third phase. |
82 | // |
83 | // Functions that don't take buffer fat pointers are modified in place. Those |
84 | // that do take such pointers have their basic blocks moved to a new function |
85 | // with arguments that are {ptr addrspace(8), i32} arguments and return values. |
86 | // This phase also records intrinsics so that they can be remangled or deleted |
87 | // later. |
88 | // |
89 | // |
90 | // ## Splitting pointer structs |
91 | // |
92 | // The meat of this pass consists of defining semantics for operations that |
93 | // produce or consume [vectors of] buffer fat pointers in terms of their |
94 | // resource and offset parts. This is accomplished throgh the `SplitPtrStructs` |
95 | // visitor. |
96 | // |
97 | // In the first pass through each function that is being lowered, the splitter |
98 | // inserts new instructions to implement the split-structures behavior, which is |
99 | // needed for correctness and performance. It records a list of "split users", |
100 | // instructions that are being replaced by operations on the resource and offset |
101 | // parts. |
102 | // |
103 | // Split users do not necessarily need to produce parts themselves ( |
104 | // a `load float, ptr addrspace(7)` does not, for example), but, if they do not |
105 | // generate fat buffer pointers, they must RAUW in their replacement |
106 | // instructions during the initial visit. |
107 | // |
108 | // When these new instructions are created, they use the split parts recorded |
109 | // for their initial arguments in order to generate their replacements, creating |
110 | // a parallel set of instructions that does not refer to the original fat |
111 | // pointer values but instead to their resource and offset components. |
112 | // |
113 | // Instructions, such as `extractvalue`, that produce buffer fat pointers from |
114 | // sources that do not have split parts, have such parts generated using |
115 | // `extractvalue`. This is also the initial handling of PHI nodes, which |
116 | // are then cleaned up. |
117 | // |
118 | // ### Conditionals |
119 | // |
120 | // PHI nodes are initially given resource parts via `extractvalue`. However, |
121 | // this is not an efficient rewrite of such nodes, as, in most cases, the |
122 | // resource part in a conditional or loop remains constant throughout the loop |
123 | // and only the offset varies. Failing to optimize away these constant resources |
124 | // would cause additional registers to be sent around loops and might lead to |
125 | // waterfall loops being generated for buffer operations due to the |
126 | // "non-uniform" resource argument. |
127 | // |
128 | // Therefore, after all instructions have been visited, the pointer splitter |
129 | // post-processes all encountered conditionals. Given a PHI node or select, |
130 | // getPossibleRsrcRoots() collects all values that the resource parts of that |
131 | // conditional's input could come from as well as collecting all conditional |
132 | // instructions encountered during the search. If, after filtering out the |
133 | // initial node itself, the set of encountered conditionals is a subset of the |
134 | // potential roots and there is a single potential resource that isn't in the |
135 | // conditional set, that value is the only possible value the resource argument |
136 | // could have throughout the control flow. |
137 | // |
138 | // If that condition is met, then a PHI node can have its resource part changed |
139 | // to the singleton value and then be replaced by a PHI on the offsets. |
140 | // Otherwise, each PHI node is split into two, one for the resource part and one |
141 | // for the offset part, which replace the temporary `extractvalue` instructions |
142 | // that were added during the first pass. |
143 | // |
144 | // Similar logic applies to `select`, where |
145 | // `%z = select i1 %cond, %cond, ptr addrspace(7) %x, ptr addrspace(7) %y` |
146 | // can be split into `%z.rsrc = %x.rsrc` and |
147 | // `%z.off = select i1 %cond, ptr i32 %x.off, i32 %y.off` |
148 | // if both `%x` and `%y` have the same resource part, but two `select` |
149 | // operations will be needed if they do not. |
150 | // |
151 | // ### Final processing |
152 | // |
153 | // After conditionals have been cleaned up, the IR for each function is |
154 | // rewritten to remove all the old instructions that have been split up. |
155 | // |
156 | // Any instruction that used to produce a buffer fat pointer (and therefore now |
157 | // produces a resource-and-offset struct after type remapping) is |
158 | // replaced as follows: |
159 | // 1. All debug value annotations are cloned to reflect that the resource part |
160 | // and offset parts are computed separately and constitute different |
161 | // fragments of the underlying source language variable. |
162 | // 2. All uses that were themselves split are replaced by a `poison` of the |
163 | // struct type, as they will themselves be erased soon. This rule, combined |
164 | // with debug handling, should leave the use lists of split instructions |
165 | // empty in almost all cases. |
166 | // 3. If a user of the original struct-valued result remains, the structure |
167 | // needed for the new types to work is constructed out of the newly-defined |
168 | // parts, and the original instruction is replaced by this structure |
169 | // before being erased. Instructions requiring this construction include |
170 | // `ret` and `insertvalue`. |
171 | // |
172 | // # Consequences |
173 | // |
174 | // This pass does not alter the CFG. |
175 | // |
176 | // Alias analysis information will become coarser, as the LLVM alias analyzer |
177 | // cannot handle the buffer intrinsics. Specifically, while we can determine |
178 | // that the following two loads do not alias: |
179 | // ``` |
180 | // %y = getelementptr i32, ptr addrspace(7) %x, i32 1 |
181 | // %a = load i32, ptr addrspace(7) %x |
182 | // %b = load i32, ptr addrspace(7) %y |
183 | // ``` |
184 | // we cannot (except through some code that runs during scheduling) determine |
185 | // that the rewritten loads below do not alias. |
186 | // ``` |
187 | // %y.off = add i32 %x.off, 1 |
188 | // %a = call @llvm.amdgcn.raw.ptr.buffer.load(ptr addrspace(8) %x.rsrc, i32 |
189 | // %x.off, ...) |
190 | // %b = call @llvm.amdgcn.raw.ptr.buffer.load(ptr addrspace(8) |
191 | // %x.rsrc, i32 %y.off, ...) |
192 | // ``` |
193 | // However, existing alias information is preserved. |
194 | //===----------------------------------------------------------------------===// |
195 | |
196 | #include "AMDGPU.h" |
197 | #include "AMDGPUTargetMachine.h" |
198 | #include "GCNSubtarget.h" |
199 | #include "SIDefines.h" |
200 | #include "llvm/ADT/SetOperations.h" |
201 | #include "llvm/ADT/SmallVector.h" |
202 | #include "llvm/Analysis/ConstantFolding.h" |
203 | #include "llvm/Analysis/Utils/Local.h" |
204 | #include "llvm/CodeGen/TargetPassConfig.h" |
205 | #include "llvm/IR/AttributeMask.h" |
206 | #include "llvm/IR/Constants.h" |
207 | #include "llvm/IR/DebugInfo.h" |
208 | #include "llvm/IR/DerivedTypes.h" |
209 | #include "llvm/IR/IRBuilder.h" |
210 | #include "llvm/IR/InstIterator.h" |
211 | #include "llvm/IR/InstVisitor.h" |
212 | #include "llvm/IR/Instructions.h" |
213 | #include "llvm/IR/Intrinsics.h" |
214 | #include "llvm/IR/IntrinsicsAMDGPU.h" |
215 | #include "llvm/IR/Metadata.h" |
216 | #include "llvm/IR/Operator.h" |
217 | #include "llvm/IR/PatternMatch.h" |
218 | #include "llvm/IR/ReplaceConstant.h" |
219 | #include "llvm/InitializePasses.h" |
220 | #include "llvm/Pass.h" |
221 | #include "llvm/Support/AtomicOrdering.h" |
222 | #include "llvm/Support/Debug.h" |
223 | #include "llvm/Support/ErrorHandling.h" |
224 | #include "llvm/Transforms/Utils/Cloning.h" |
225 | #include "llvm/Transforms/Utils/Local.h" |
226 | #include "llvm/Transforms/Utils/ValueMapper.h" |
227 | |
228 | #define DEBUG_TYPE "amdgpu-lower-buffer-fat-pointers" |
229 | |
230 | using namespace llvm; |
231 | |
232 | static constexpr unsigned BufferOffsetWidth = 32; |
233 | |
234 | namespace { |
235 | /// Recursively replace instances of ptr addrspace(7) and vector<Nxptr |
236 | /// addrspace(7)> with some other type as defined by the relevant subclass. |
237 | class BufferFatPtrTypeLoweringBase : public ValueMapTypeRemapper { |
238 | DenseMap<Type *, Type *> Map; |
239 | |
240 | Type *remapTypeImpl(Type *Ty, SmallPtrSetImpl<StructType *> &Seen); |
241 | |
242 | protected: |
243 | virtual Type *remapScalar(PointerType *PT) = 0; |
244 | virtual Type *remapVector(VectorType *VT) = 0; |
245 | |
246 | const DataLayout &DL; |
247 | |
248 | public: |
249 | BufferFatPtrTypeLoweringBase(const DataLayout &DL) : DL(DL) {} |
250 | Type *remapType(Type *SrcTy) override; |
251 | void clear() { Map.clear(); } |
252 | }; |
253 | |
254 | /// Remap ptr addrspace(7) to i160 and vector<Nxptr addrspace(7)> to |
255 | /// vector<Nxi60> in order to correctly handling loading/storing these values |
256 | /// from memory. |
257 | class BufferFatPtrToIntTypeMap : public BufferFatPtrTypeLoweringBase { |
258 | using BufferFatPtrTypeLoweringBase::BufferFatPtrTypeLoweringBase; |
259 | |
260 | protected: |
261 | Type *remapScalar(PointerType *PT) override { return DL.getIntPtrType(PT); } |
262 | Type *remapVector(VectorType *VT) override { return DL.getIntPtrType(VT); } |
263 | }; |
264 | |
265 | /// Remap ptr addrspace(7) to {ptr addrspace(8), i32} (the resource and offset |
266 | /// parts of the pointer) so that we can easily rewrite operations on these |
267 | /// values that aren't loading them from or storing them to memory. |
268 | class BufferFatPtrToStructTypeMap : public BufferFatPtrTypeLoweringBase { |
269 | using BufferFatPtrTypeLoweringBase::BufferFatPtrTypeLoweringBase; |
270 | |
271 | protected: |
272 | Type *remapScalar(PointerType *PT) override; |
273 | Type *remapVector(VectorType *VT) override; |
274 | }; |
275 | } // namespace |
276 | |
277 | // This code is adapted from the type remapper in lib/Linker/IRMover.cpp |
278 | Type *BufferFatPtrTypeLoweringBase::remapTypeImpl( |
279 | Type *Ty, SmallPtrSetImpl<StructType *> &Seen) { |
280 | Type **Entry = &Map[Ty]; |
281 | if (*Entry) |
282 | return *Entry; |
283 | if (auto *PT = dyn_cast<PointerType>(Val: Ty)) { |
284 | if (PT->getAddressSpace() == AMDGPUAS::BUFFER_FAT_POINTER) { |
285 | return *Entry = remapScalar(PT); |
286 | } |
287 | } |
288 | if (auto *VT = dyn_cast<VectorType>(Val: Ty)) { |
289 | auto *PT = dyn_cast<PointerType>(Val: VT->getElementType()); |
290 | if (PT && PT->getAddressSpace() == AMDGPUAS::BUFFER_FAT_POINTER) { |
291 | return *Entry = remapVector(VT); |
292 | } |
293 | return *Entry = Ty; |
294 | } |
295 | // Whether the type is one that is structurally uniqued - that is, if it is |
296 | // not a named struct (the only kind of type where multiple structurally |
297 | // identical types that have a distinct `Type*`) |
298 | StructType *TyAsStruct = dyn_cast<StructType>(Val: Ty); |
299 | bool IsUniqued = !TyAsStruct || TyAsStruct->isLiteral(); |
300 | // Base case for ints, floats, opaque pointers, and so on, which don't |
301 | // require recursion. |
302 | if (Ty->getNumContainedTypes() == 0 && IsUniqued) |
303 | return *Entry = Ty; |
304 | if (!IsUniqued) { |
305 | // Create a dummy type for recursion purposes. |
306 | if (!Seen.insert(Ptr: TyAsStruct).second) { |
307 | StructType *Placeholder = StructType::create(Context&: Ty->getContext()); |
308 | return *Entry = Placeholder; |
309 | } |
310 | } |
311 | bool Changed = false; |
312 | SmallVector<Type *> ElementTypes(Ty->getNumContainedTypes(), nullptr); |
313 | for (unsigned int I = 0, E = Ty->getNumContainedTypes(); I < E; ++I) { |
314 | Type *OldElem = Ty->getContainedType(i: I); |
315 | Type *NewElem = remapTypeImpl(Ty: OldElem, Seen); |
316 | ElementTypes[I] = NewElem; |
317 | Changed |= (OldElem != NewElem); |
318 | } |
319 | // Recursive calls to remapTypeImpl() may have invalidated pointer. |
320 | Entry = &Map[Ty]; |
321 | if (!Changed) { |
322 | return *Entry = Ty; |
323 | } |
324 | if (auto *ArrTy = dyn_cast<ArrayType>(Val: Ty)) |
325 | return *Entry = ArrayType::get(ElementType: ElementTypes[0], NumElements: ArrTy->getNumElements()); |
326 | if (auto *FnTy = dyn_cast<FunctionType>(Val: Ty)) |
327 | return *Entry = FunctionType::get(Result: ElementTypes[0], |
328 | Params: ArrayRef(ElementTypes).slice(N: 1), |
329 | isVarArg: FnTy->isVarArg()); |
330 | if (auto *STy = dyn_cast<StructType>(Val: Ty)) { |
331 | // Genuine opaque types don't have a remapping. |
332 | if (STy->isOpaque()) |
333 | return *Entry = Ty; |
334 | bool IsPacked = STy->isPacked(); |
335 | if (IsUniqued) |
336 | return *Entry = StructType::get(Context&: Ty->getContext(), Elements: ElementTypes, isPacked: IsPacked); |
337 | SmallString<16> Name(STy->getName()); |
338 | STy->setName("" ); |
339 | Type **RecursionEntry = &Map[Ty]; |
340 | if (*RecursionEntry) { |
341 | auto *Placeholder = cast<StructType>(Val: *RecursionEntry); |
342 | Placeholder->setBody(Elements: ElementTypes, isPacked: IsPacked); |
343 | Placeholder->setName(Name); |
344 | return *Entry = Placeholder; |
345 | } |
346 | return *Entry = StructType::create(Context&: Ty->getContext(), Elements: ElementTypes, Name, |
347 | isPacked: IsPacked); |
348 | } |
349 | llvm_unreachable("Unknown type of type that contains elements" ); |
350 | } |
351 | |
352 | Type *BufferFatPtrTypeLoweringBase::remapType(Type *SrcTy) { |
353 | SmallPtrSet<StructType *, 2> Visited; |
354 | return remapTypeImpl(Ty: SrcTy, Seen&: Visited); |
355 | } |
356 | |
357 | Type *BufferFatPtrToStructTypeMap::remapScalar(PointerType *PT) { |
358 | LLVMContext &Ctx = PT->getContext(); |
359 | return StructType::get(elt1: PointerType::get(C&: Ctx, AddressSpace: AMDGPUAS::BUFFER_RESOURCE), |
360 | elts: IntegerType::get(C&: Ctx, NumBits: BufferOffsetWidth)); |
361 | } |
362 | |
363 | Type *BufferFatPtrToStructTypeMap::remapVector(VectorType *VT) { |
364 | ElementCount EC = VT->getElementCount(); |
365 | LLVMContext &Ctx = VT->getContext(); |
366 | Type *RsrcVec = |
367 | VectorType::get(ElementType: PointerType::get(C&: Ctx, AddressSpace: AMDGPUAS::BUFFER_RESOURCE), EC); |
368 | Type *OffVec = VectorType::get(ElementType: IntegerType::get(C&: Ctx, NumBits: BufferOffsetWidth), EC); |
369 | return StructType::get(elt1: RsrcVec, elts: OffVec); |
370 | } |
371 | |
372 | static bool isBufferFatPtrOrVector(Type *Ty) { |
373 | if (auto *PT = dyn_cast<PointerType>(Val: Ty->getScalarType())) |
374 | return PT->getAddressSpace() == AMDGPUAS::BUFFER_FAT_POINTER; |
375 | return false; |
376 | } |
377 | |
378 | // True if the type is {ptr addrspace(8), i32} or a struct containing vectors of |
379 | // those types. Used to quickly skip instructions we don't need to process. |
380 | static bool isSplitFatPtr(Type *Ty) { |
381 | auto *ST = dyn_cast<StructType>(Val: Ty); |
382 | if (!ST) |
383 | return false; |
384 | if (!ST->isLiteral() || ST->getNumElements() != 2) |
385 | return false; |
386 | auto *MaybeRsrc = |
387 | dyn_cast<PointerType>(Val: ST->getElementType(N: 0)->getScalarType()); |
388 | auto *MaybeOff = |
389 | dyn_cast<IntegerType>(Val: ST->getElementType(N: 1)->getScalarType()); |
390 | return MaybeRsrc && MaybeOff && |
391 | MaybeRsrc->getAddressSpace() == AMDGPUAS::BUFFER_RESOURCE && |
392 | MaybeOff->getBitWidth() == BufferOffsetWidth; |
393 | } |
394 | |
395 | // True if the result type or any argument types are buffer fat pointers. |
396 | static bool isBufferFatPtrConst(Constant *C) { |
397 | Type *T = C->getType(); |
398 | return isBufferFatPtrOrVector(Ty: T) || any_of(Range: C->operands(), P: [](const Use &U) { |
399 | return isBufferFatPtrOrVector(Ty: U.get()->getType()); |
400 | }); |
401 | } |
402 | |
403 | namespace { |
404 | /// Convert [vectors of] buffer fat pointers to integers when they are read from |
405 | /// or stored to memory. This ensures that these pointers will have the same |
406 | /// memory layout as before they are lowered, even though they will no longer |
407 | /// have their previous layout in registers/in the program (they'll be broken |
408 | /// down into resource and offset parts). This has the downside of imposing |
409 | /// marshalling costs when reading or storing these values, but since placing |
410 | /// such pointers into memory is an uncommon operation at best, we feel that |
411 | /// this cost is acceptable for better performance in the common case. |
412 | class StoreFatPtrsAsIntsVisitor |
413 | : public InstVisitor<StoreFatPtrsAsIntsVisitor, bool> { |
414 | BufferFatPtrToIntTypeMap *TypeMap; |
415 | |
416 | ValueToValueMapTy ConvertedForStore; |
417 | |
418 | IRBuilder<> IRB; |
419 | |
420 | // Convert all the buffer fat pointers within the input value to inttegers |
421 | // so that it can be stored in memory. |
422 | Value *fatPtrsToInts(Value *V, Type *From, Type *To, const Twine &Name); |
423 | // Convert all the i160s that need to be buffer fat pointers (as specified) |
424 | // by the To type) into those pointers to preserve the semantics of the rest |
425 | // of the program. |
426 | Value *intsToFatPtrs(Value *V, Type *From, Type *To, const Twine &Name); |
427 | |
428 | public: |
429 | StoreFatPtrsAsIntsVisitor(BufferFatPtrToIntTypeMap *TypeMap, LLVMContext &Ctx) |
430 | : TypeMap(TypeMap), IRB(Ctx) {} |
431 | bool processFunction(Function &F); |
432 | |
433 | bool visitInstruction(Instruction &I) { return false; } |
434 | bool visitAllocaInst(AllocaInst &I); |
435 | bool visitLoadInst(LoadInst &LI); |
436 | bool visitStoreInst(StoreInst &SI); |
437 | bool visitGetElementPtrInst(GetElementPtrInst &I); |
438 | }; |
439 | } // namespace |
440 | |
441 | Value *StoreFatPtrsAsIntsVisitor::fatPtrsToInts(Value *V, Type *From, Type *To, |
442 | const Twine &Name) { |
443 | if (From == To) |
444 | return V; |
445 | ValueToValueMapTy::iterator Find = ConvertedForStore.find(Val: V); |
446 | if (Find != ConvertedForStore.end()) |
447 | return Find->second; |
448 | if (isBufferFatPtrOrVector(Ty: From)) { |
449 | Value *Cast = IRB.CreatePtrToInt(V, DestTy: To, Name: Name + ".int" ); |
450 | ConvertedForStore[V] = Cast; |
451 | return Cast; |
452 | } |
453 | if (From->getNumContainedTypes() == 0) |
454 | return V; |
455 | // Structs, arrays, and other compound types. |
456 | Value *Ret = PoisonValue::get(T: To); |
457 | if (auto *AT = dyn_cast<ArrayType>(Val: From)) { |
458 | Type *FromPart = AT->getArrayElementType(); |
459 | Type *ToPart = cast<ArrayType>(Val: To)->getElementType(); |
460 | for (uint64_t I = 0, E = AT->getArrayNumElements(); I < E; ++I) { |
461 | Value *Field = IRB.CreateExtractValue(Agg: V, Idxs: I); |
462 | Value *NewField = |
463 | fatPtrsToInts(V: Field, From: FromPart, To: ToPart, Name: Name + "." + Twine(I)); |
464 | Ret = IRB.CreateInsertValue(Agg: Ret, Val: NewField, Idxs: I); |
465 | } |
466 | } else { |
467 | for (auto [Idx, FromPart, ToPart] : |
468 | enumerate(First: From->subtypes(), Rest: To->subtypes())) { |
469 | Value *Field = IRB.CreateExtractValue(Agg: V, Idxs: Idx); |
470 | Value *NewField = |
471 | fatPtrsToInts(V: Field, From: FromPart, To: ToPart, Name: Name + "." + Twine(Idx)); |
472 | Ret = IRB.CreateInsertValue(Agg: Ret, Val: NewField, Idxs: Idx); |
473 | } |
474 | } |
475 | ConvertedForStore[V] = Ret; |
476 | return Ret; |
477 | } |
478 | |
479 | Value *StoreFatPtrsAsIntsVisitor::intsToFatPtrs(Value *V, Type *From, Type *To, |
480 | const Twine &Name) { |
481 | if (From == To) |
482 | return V; |
483 | if (isBufferFatPtrOrVector(Ty: To)) { |
484 | Value *Cast = IRB.CreateIntToPtr(V, DestTy: To, Name: Name + ".ptr" ); |
485 | return Cast; |
486 | } |
487 | if (From->getNumContainedTypes() == 0) |
488 | return V; |
489 | // Structs, arrays, and other compound types. |
490 | Value *Ret = PoisonValue::get(T: To); |
491 | if (auto *AT = dyn_cast<ArrayType>(Val: From)) { |
492 | Type *FromPart = AT->getArrayElementType(); |
493 | Type *ToPart = cast<ArrayType>(Val: To)->getElementType(); |
494 | for (uint64_t I = 0, E = AT->getArrayNumElements(); I < E; ++I) { |
495 | Value *Field = IRB.CreateExtractValue(Agg: V, Idxs: I); |
496 | Value *NewField = |
497 | intsToFatPtrs(V: Field, From: FromPart, To: ToPart, Name: Name + "." + Twine(I)); |
498 | Ret = IRB.CreateInsertValue(Agg: Ret, Val: NewField, Idxs: I); |
499 | } |
500 | } else { |
501 | for (auto [Idx, FromPart, ToPart] : |
502 | enumerate(First: From->subtypes(), Rest: To->subtypes())) { |
503 | Value *Field = IRB.CreateExtractValue(Agg: V, Idxs: Idx); |
504 | Value *NewField = |
505 | intsToFatPtrs(V: Field, From: FromPart, To: ToPart, Name: Name + "." + Twine(Idx)); |
506 | Ret = IRB.CreateInsertValue(Agg: Ret, Val: NewField, Idxs: Idx); |
507 | } |
508 | } |
509 | return Ret; |
510 | } |
511 | |
512 | bool StoreFatPtrsAsIntsVisitor::processFunction(Function &F) { |
513 | bool Changed = false; |
514 | // The visitors will mutate GEPs and allocas, but will push loads and stores |
515 | // to the worklist to avoid invalidation. |
516 | for (Instruction &I : make_early_inc_range(Range: instructions(F))) { |
517 | Changed |= visit(I); |
518 | } |
519 | ConvertedForStore.clear(); |
520 | return Changed; |
521 | } |
522 | |
523 | bool StoreFatPtrsAsIntsVisitor::visitAllocaInst(AllocaInst &I) { |
524 | Type *Ty = I.getAllocatedType(); |
525 | Type *NewTy = TypeMap->remapType(SrcTy: Ty); |
526 | if (Ty == NewTy) |
527 | return false; |
528 | I.setAllocatedType(NewTy); |
529 | return true; |
530 | } |
531 | |
532 | bool StoreFatPtrsAsIntsVisitor::visitGetElementPtrInst(GetElementPtrInst &I) { |
533 | Type *Ty = I.getSourceElementType(); |
534 | Type *NewTy = TypeMap->remapType(SrcTy: Ty); |
535 | if (Ty == NewTy) |
536 | return false; |
537 | // We'll be rewriting the type `ptr addrspace(7)` out of existence soon, so |
538 | // make sure GEPs don't have different semantics with the new type. |
539 | I.setSourceElementType(NewTy); |
540 | I.setResultElementType(TypeMap->remapType(SrcTy: I.getResultElementType())); |
541 | return true; |
542 | } |
543 | |
544 | bool StoreFatPtrsAsIntsVisitor::visitLoadInst(LoadInst &LI) { |
545 | Type *Ty = LI.getType(); |
546 | Type *IntTy = TypeMap->remapType(SrcTy: Ty); |
547 | if (Ty == IntTy) |
548 | return false; |
549 | |
550 | IRB.SetInsertPoint(&LI); |
551 | auto *NLI = cast<LoadInst>(Val: LI.clone()); |
552 | NLI->mutateType(Ty: IntTy); |
553 | NLI = IRB.Insert(I: NLI); |
554 | copyMetadataForLoad(Dest&: *NLI, Source: LI); |
555 | NLI->takeName(V: &LI); |
556 | |
557 | Value *CastBack = intsToFatPtrs(V: NLI, From: IntTy, To: Ty, Name: NLI->getName()); |
558 | LI.replaceAllUsesWith(V: CastBack); |
559 | LI.eraseFromParent(); |
560 | return true; |
561 | } |
562 | |
563 | bool StoreFatPtrsAsIntsVisitor::visitStoreInst(StoreInst &SI) { |
564 | Value *V = SI.getValueOperand(); |
565 | Type *Ty = V->getType(); |
566 | Type *IntTy = TypeMap->remapType(SrcTy: Ty); |
567 | if (Ty == IntTy) |
568 | return false; |
569 | |
570 | IRB.SetInsertPoint(&SI); |
571 | Value *IntV = fatPtrsToInts(V, From: Ty, To: IntTy, Name: V->getName()); |
572 | for (auto *Dbg : at::getAssignmentMarkers(Inst: &SI)) |
573 | Dbg->setValue(IntV); |
574 | |
575 | SI.setOperand(i_nocapture: 0, Val_nocapture: IntV); |
576 | return true; |
577 | } |
578 | |
579 | /// Return the ptr addrspace(8) and i32 (resource and offset parts) in a lowered |
580 | /// buffer fat pointer constant. |
581 | static std::pair<Constant *, Constant *> |
582 | splitLoweredFatBufferConst(Constant *C) { |
583 | assert(isSplitFatPtr(C->getType()) && "Not a split fat buffer pointer" ); |
584 | return std::make_pair(x: C->getAggregateElement(Elt: 0u), y: C->getAggregateElement(Elt: 1u)); |
585 | } |
586 | |
587 | namespace { |
588 | /// Handle the remapping of ptr addrspace(7) constants. |
589 | class FatPtrConstMaterializer final : public ValueMaterializer { |
590 | BufferFatPtrToStructTypeMap *TypeMap; |
591 | // An internal mapper that is used to recurse into the arguments of constants. |
592 | // While the documentation for `ValueMapper` specifies not to use it |
593 | // recursively, examination of the logic in mapValue() shows that it can |
594 | // safely be used recursively when handling constants, like it does in its own |
595 | // logic. |
596 | ValueMapper InternalMapper; |
597 | |
598 | Constant *materializeBufferFatPtrConst(Constant *C); |
599 | |
600 | public: |
601 | // UnderlyingMap is the value map this materializer will be filling. |
602 | FatPtrConstMaterializer(BufferFatPtrToStructTypeMap *TypeMap, |
603 | ValueToValueMapTy &UnderlyingMap) |
604 | : TypeMap(TypeMap), |
605 | InternalMapper(UnderlyingMap, RF_None, TypeMap, this) {} |
606 | virtual ~FatPtrConstMaterializer() = default; |
607 | |
608 | Value *materialize(Value *V) override; |
609 | }; |
610 | } // namespace |
611 | |
612 | Constant *FatPtrConstMaterializer::materializeBufferFatPtrConst(Constant *C) { |
613 | Type *SrcTy = C->getType(); |
614 | auto *NewTy = dyn_cast<StructType>(Val: TypeMap->remapType(SrcTy)); |
615 | if (C->isNullValue()) |
616 | return ConstantAggregateZero::getNullValue(Ty: NewTy); |
617 | if (isa<PoisonValue>(Val: C)) { |
618 | return ConstantStruct::get(T: NewTy, |
619 | V: {PoisonValue::get(T: NewTy->getElementType(N: 0)), |
620 | PoisonValue::get(T: NewTy->getElementType(N: 1))}); |
621 | } |
622 | if (isa<UndefValue>(Val: C)) { |
623 | return ConstantStruct::get(T: NewTy, |
624 | V: {UndefValue::get(T: NewTy->getElementType(N: 0)), |
625 | UndefValue::get(T: NewTy->getElementType(N: 1))}); |
626 | } |
627 | |
628 | if (auto *VC = dyn_cast<ConstantVector>(Val: C)) { |
629 | if (Constant *S = VC->getSplatValue()) { |
630 | Constant *NewS = InternalMapper.mapConstant(C: *S); |
631 | if (!NewS) |
632 | return nullptr; |
633 | auto [Rsrc, Off] = splitLoweredFatBufferConst(C: NewS); |
634 | auto EC = VC->getType()->getElementCount(); |
635 | return ConstantStruct::get(T: NewTy, V: {ConstantVector::getSplat(EC, Elt: Rsrc), |
636 | ConstantVector::getSplat(EC, Elt: Off)}); |
637 | } |
638 | SmallVector<Constant *> Rsrcs; |
639 | SmallVector<Constant *> Offs; |
640 | for (Value *Op : VC->operand_values()) { |
641 | auto *NewOp = dyn_cast_or_null<Constant>(Val: InternalMapper.mapValue(V: *Op)); |
642 | if (!NewOp) |
643 | return nullptr; |
644 | auto [Rsrc, Off] = splitLoweredFatBufferConst(C: NewOp); |
645 | Rsrcs.push_back(Elt: Rsrc); |
646 | Offs.push_back(Elt: Off); |
647 | } |
648 | Constant *RsrcVec = ConstantVector::get(V: Rsrcs); |
649 | Constant *OffVec = ConstantVector::get(V: Offs); |
650 | return ConstantStruct::get(T: NewTy, V: {RsrcVec, OffVec}); |
651 | } |
652 | |
653 | if (isa<GlobalValue>(Val: C)) |
654 | report_fatal_error(reason: "Global values containing ptr addrspace(7) (buffer " |
655 | "fat pointer) values are not supported" ); |
656 | |
657 | if (isa<ConstantExpr>(Val: C)) |
658 | report_fatal_error(reason: "Constant exprs containing ptr addrspace(7) (buffer " |
659 | "fat pointer) values should have been expanded earlier" ); |
660 | |
661 | return nullptr; |
662 | } |
663 | |
664 | Value *FatPtrConstMaterializer::materialize(Value *V) { |
665 | Constant *C = dyn_cast<Constant>(Val: V); |
666 | if (!C) |
667 | return nullptr; |
668 | // Structs and other types that happen to contain fat pointers get remapped |
669 | // by the mapValue() logic. |
670 | if (!isBufferFatPtrConst(C)) |
671 | return nullptr; |
672 | return materializeBufferFatPtrConst(C); |
673 | } |
674 | |
675 | using PtrParts = std::pair<Value *, Value *>; |
676 | namespace { |
677 | // The visitor returns the resource and offset parts for an instruction if they |
678 | // can be computed, or (nullptr, nullptr) for cases that don't have a meaningful |
679 | // value mapping. |
680 | class SplitPtrStructs : public InstVisitor<SplitPtrStructs, PtrParts> { |
681 | ValueToValueMapTy RsrcParts; |
682 | ValueToValueMapTy OffParts; |
683 | |
684 | // Track instructions that have been rewritten into a user of the component |
685 | // parts of their ptr addrspace(7) input. Instructions that produced |
686 | // ptr addrspace(7) parts should **not** be RAUW'd before being added to this |
687 | // set, as that replacement will be handled in a post-visit step. However, |
688 | // instructions that yield values that aren't fat pointers (ex. ptrtoint) |
689 | // should RAUW themselves with new instructions that use the split parts |
690 | // of their arguments during processing. |
691 | DenseSet<Instruction *> SplitUsers; |
692 | |
693 | // Nodes that need a second look once we've computed the parts for all other |
694 | // instructions to see if, for example, we really need to phi on the resource |
695 | // part. |
696 | SmallVector<Instruction *> Conditionals; |
697 | // Temporary instructions produced while lowering conditionals that should be |
698 | // killed. |
699 | SmallVector<Instruction *> ConditionalTemps; |
700 | |
701 | // Subtarget info, needed for determining what cache control bits to set. |
702 | const TargetMachine *TM; |
703 | const GCNSubtarget *ST = nullptr; |
704 | |
705 | IRBuilder<> IRB; |
706 | |
707 | // Copy metadata between instructions if applicable. |
708 | void copyMetadata(Value *Dest, Value *Src); |
709 | |
710 | // Get the resource and offset parts of the value V, inserting appropriate |
711 | // extractvalue calls if needed. |
712 | PtrParts getPtrParts(Value *V); |
713 | |
714 | // Given an instruction that could produce multiple resource parts (a PHI or |
715 | // select), collect the set of possible instructions that could have provided |
716 | // its resource parts that it could have (the `Roots`) and the set of |
717 | // conditional instructions visited during the search (`Seen`). If, after |
718 | // removing the root of the search from `Seen` and `Roots`, `Seen` is a subset |
719 | // of `Roots` and `Roots - Seen` contains one element, the resource part of |
720 | // that element can replace the resource part of all other elements in `Seen`. |
721 | void getPossibleRsrcRoots(Instruction *I, SmallPtrSetImpl<Value *> &Roots, |
722 | SmallPtrSetImpl<Value *> &Seen); |
723 | void processConditionals(); |
724 | |
725 | // If an instruction hav been split into resource and offset parts, |
726 | // delete that instruction. If any of its uses have not themselves been split |
727 | // into parts (for example, an insertvalue), construct the structure |
728 | // that the type rewrites declared should be produced by the dying instruction |
729 | // and use that. |
730 | // Also, kill the temporary extractvalue operations produced by the two-stage |
731 | // lowering of PHIs and conditionals. |
732 | void killAndReplaceSplitInstructions(SmallVectorImpl<Instruction *> &Origs); |
733 | |
734 | void setAlign(CallInst *Intr, Align A, unsigned RsrcArgIdx); |
735 | void insertPreMemOpFence(AtomicOrdering Order, SyncScope::ID SSID); |
736 | void insertPostMemOpFence(AtomicOrdering Order, SyncScope::ID SSID); |
737 | Value *handleMemoryInst(Instruction *I, Value *Arg, Value *Ptr, Type *Ty, |
738 | Align Alignment, AtomicOrdering Order, |
739 | bool IsVolatile, SyncScope::ID SSID); |
740 | |
741 | public: |
742 | SplitPtrStructs(LLVMContext &Ctx, const TargetMachine *TM) |
743 | : TM(TM), IRB(Ctx) {} |
744 | |
745 | void processFunction(Function &F); |
746 | |
747 | PtrParts visitInstruction(Instruction &I); |
748 | PtrParts visitLoadInst(LoadInst &LI); |
749 | PtrParts visitStoreInst(StoreInst &SI); |
750 | PtrParts visitAtomicRMWInst(AtomicRMWInst &AI); |
751 | PtrParts visitAtomicCmpXchgInst(AtomicCmpXchgInst &AI); |
752 | PtrParts visitGetElementPtrInst(GetElementPtrInst &GEP); |
753 | |
754 | PtrParts visitPtrToIntInst(PtrToIntInst &PI); |
755 | PtrParts visitIntToPtrInst(IntToPtrInst &IP); |
756 | PtrParts visitAddrSpaceCastInst(AddrSpaceCastInst &I); |
757 | PtrParts visitICmpInst(ICmpInst &Cmp); |
758 | PtrParts visitFreezeInst(FreezeInst &I); |
759 | |
760 | PtrParts visitExtractElementInst(ExtractElementInst &I); |
761 | PtrParts visitInsertElementInst(InsertElementInst &I); |
762 | PtrParts visitShuffleVectorInst(ShuffleVectorInst &I); |
763 | |
764 | PtrParts visitPHINode(PHINode &PHI); |
765 | PtrParts visitSelectInst(SelectInst &SI); |
766 | |
767 | PtrParts visitIntrinsicInst(IntrinsicInst &II); |
768 | }; |
769 | } // namespace |
770 | |
771 | void SplitPtrStructs::copyMetadata(Value *Dest, Value *Src) { |
772 | auto *DestI = dyn_cast<Instruction>(Val: Dest); |
773 | auto *SrcI = dyn_cast<Instruction>(Val: Src); |
774 | |
775 | if (!DestI || !SrcI) |
776 | return; |
777 | |
778 | DestI->copyMetadata(SrcInst: *SrcI); |
779 | } |
780 | |
781 | PtrParts SplitPtrStructs::getPtrParts(Value *V) { |
782 | assert(isSplitFatPtr(V->getType()) && "it's not meaningful to get the parts " |
783 | "of something that wasn't rewritten" ); |
784 | auto *RsrcEntry = &RsrcParts[V]; |
785 | auto *OffEntry = &OffParts[V]; |
786 | if (*RsrcEntry && *OffEntry) |
787 | return {*RsrcEntry, *OffEntry}; |
788 | |
789 | if (auto *C = dyn_cast<Constant>(Val: V)) { |
790 | auto [Rsrc, Off] = splitLoweredFatBufferConst(C); |
791 | return {*RsrcEntry = Rsrc, *OffEntry = Off}; |
792 | } |
793 | |
794 | IRBuilder<>::InsertPointGuard Guard(IRB); |
795 | if (auto *I = dyn_cast<Instruction>(Val: V)) { |
796 | LLVM_DEBUG(dbgs() << "Recursing to split parts of " << *I << "\n" ); |
797 | auto [Rsrc, Off] = visit(I&: *I); |
798 | if (Rsrc && Off) |
799 | return {*RsrcEntry = Rsrc, *OffEntry = Off}; |
800 | // We'll be creating the new values after the relevant instruction. |
801 | // This instruction generates a value and so isn't a terminator. |
802 | IRB.SetInsertPoint(*I->getInsertionPointAfterDef()); |
803 | IRB.SetCurrentDebugLocation(I->getDebugLoc()); |
804 | } else if (auto *A = dyn_cast<Argument>(Val: V)) { |
805 | IRB.SetInsertPointPastAllocas(A->getParent()); |
806 | IRB.SetCurrentDebugLocation(DebugLoc()); |
807 | } |
808 | Value *Rsrc = IRB.CreateExtractValue(Agg: V, Idxs: 0, Name: V->getName() + ".rsrc" ); |
809 | Value *Off = IRB.CreateExtractValue(Agg: V, Idxs: 1, Name: V->getName() + ".off" ); |
810 | return {*RsrcEntry = Rsrc, *OffEntry = Off}; |
811 | } |
812 | |
813 | /// Returns the instruction that defines the resource part of the value V. |
814 | /// Note that this is not getUnderlyingObject(), since that looks through |
815 | /// operations like ptrmask which might modify the resource part. |
816 | /// |
817 | /// We can limit ourselves to just looking through GEPs followed by looking |
818 | /// through addrspacecasts because only those two operations preserve the |
819 | /// resource part, and because operations on an `addrspace(8)` (which is the |
820 | /// legal input to this addrspacecast) would produce a different resource part. |
821 | static Value *rsrcPartRoot(Value *V) { |
822 | while (auto *GEP = dyn_cast<GEPOperator>(Val: V)) |
823 | V = GEP->getPointerOperand(); |
824 | while (auto *ASC = dyn_cast<AddrSpaceCastOperator>(Val: V)) |
825 | V = ASC->getPointerOperand(); |
826 | return V; |
827 | } |
828 | |
829 | void SplitPtrStructs::getPossibleRsrcRoots(Instruction *I, |
830 | SmallPtrSetImpl<Value *> &Roots, |
831 | SmallPtrSetImpl<Value *> &Seen) { |
832 | if (auto *PHI = dyn_cast<PHINode>(Val: I)) { |
833 | if (!Seen.insert(Ptr: I).second) |
834 | return; |
835 | for (Value *In : PHI->incoming_values()) { |
836 | In = rsrcPartRoot(V: In); |
837 | Roots.insert(Ptr: In); |
838 | if (isa<PHINode, SelectInst>(Val: In)) |
839 | getPossibleRsrcRoots(I: cast<Instruction>(Val: In), Roots, Seen); |
840 | } |
841 | } else if (auto *SI = dyn_cast<SelectInst>(Val: I)) { |
842 | if (!Seen.insert(Ptr: SI).second) |
843 | return; |
844 | Value *TrueVal = rsrcPartRoot(V: SI->getTrueValue()); |
845 | Value *FalseVal = rsrcPartRoot(V: SI->getFalseValue()); |
846 | Roots.insert(Ptr: TrueVal); |
847 | Roots.insert(Ptr: FalseVal); |
848 | if (isa<PHINode, SelectInst>(Val: TrueVal)) |
849 | getPossibleRsrcRoots(I: cast<Instruction>(Val: TrueVal), Roots, Seen); |
850 | if (isa<PHINode, SelectInst>(Val: FalseVal)) |
851 | getPossibleRsrcRoots(I: cast<Instruction>(Val: FalseVal), Roots, Seen); |
852 | } else { |
853 | llvm_unreachable("getPossibleRsrcParts() only works on phi and select" ); |
854 | } |
855 | } |
856 | |
857 | void SplitPtrStructs::processConditionals() { |
858 | SmallDenseMap<Instruction *, Value *> FoundRsrcs; |
859 | SmallPtrSet<Value *, 4> Roots; |
860 | SmallPtrSet<Value *, 4> Seen; |
861 | for (Instruction *I : Conditionals) { |
862 | // These have to exist by now because we've visited these nodes. |
863 | Value *Rsrc = RsrcParts[I]; |
864 | Value *Off = OffParts[I]; |
865 | assert(Rsrc && Off && "must have visited conditionals by now" ); |
866 | |
867 | std::optional<Value *> MaybeRsrc; |
868 | auto MaybeFoundRsrc = FoundRsrcs.find(Val: I); |
869 | if (MaybeFoundRsrc != FoundRsrcs.end()) { |
870 | MaybeRsrc = MaybeFoundRsrc->second; |
871 | } else { |
872 | IRBuilder<>::InsertPointGuard Guard(IRB); |
873 | Roots.clear(); |
874 | Seen.clear(); |
875 | getPossibleRsrcRoots(I, Roots, Seen); |
876 | LLVM_DEBUG(dbgs() << "Processing conditional: " << *I << "\n" ); |
877 | #ifndef NDEBUG |
878 | for (Value *V : Roots) |
879 | LLVM_DEBUG(dbgs() << "Root: " << *V << "\n" ); |
880 | for (Value *V : Seen) |
881 | LLVM_DEBUG(dbgs() << "Seen: " << *V << "\n" ); |
882 | #endif |
883 | // If we are our own possible root, then we shouldn't block our |
884 | // replacement with a valid incoming value. |
885 | Roots.erase(Ptr: I); |
886 | // We don't want to block the optimization for conditionals that don't |
887 | // refer to themselves but did see themselves during the traversal. |
888 | Seen.erase(Ptr: I); |
889 | |
890 | if (set_is_subset(S1: Seen, S2: Roots)) { |
891 | auto Diff = set_difference(S1: Roots, S2: Seen); |
892 | if (Diff.size() == 1) { |
893 | Value *RootVal = *Diff.begin(); |
894 | // Handle the case where previous loops already looked through |
895 | // an addrspacecast. |
896 | if (isSplitFatPtr(Ty: RootVal->getType())) |
897 | MaybeRsrc = std::get<0>(in: getPtrParts(V: RootVal)); |
898 | else |
899 | MaybeRsrc = RootVal; |
900 | } |
901 | } |
902 | } |
903 | |
904 | if (auto *PHI = dyn_cast<PHINode>(Val: I)) { |
905 | Value *NewRsrc; |
906 | StructType *PHITy = cast<StructType>(Val: PHI->getType()); |
907 | IRB.SetInsertPoint(*PHI->getInsertionPointAfterDef()); |
908 | IRB.SetCurrentDebugLocation(PHI->getDebugLoc()); |
909 | if (MaybeRsrc) { |
910 | NewRsrc = *MaybeRsrc; |
911 | } else { |
912 | Type *RsrcTy = PHITy->getElementType(N: 0); |
913 | auto *RsrcPHI = IRB.CreatePHI(Ty: RsrcTy, NumReservedValues: PHI->getNumIncomingValues()); |
914 | RsrcPHI->takeName(V: Rsrc); |
915 | for (auto [V, BB] : llvm::zip(t: PHI->incoming_values(), u: PHI->blocks())) { |
916 | Value *VRsrc = std::get<0>(in: getPtrParts(V)); |
917 | RsrcPHI->addIncoming(V: VRsrc, BB); |
918 | } |
919 | copyMetadata(Dest: RsrcPHI, Src: PHI); |
920 | NewRsrc = RsrcPHI; |
921 | } |
922 | |
923 | Type *OffTy = PHITy->getElementType(N: 1); |
924 | auto *NewOff = IRB.CreatePHI(Ty: OffTy, NumReservedValues: PHI->getNumIncomingValues()); |
925 | NewOff->takeName(V: Off); |
926 | for (auto [V, BB] : llvm::zip(t: PHI->incoming_values(), u: PHI->blocks())) { |
927 | assert(OffParts.count(V) && "An offset part had to be created by now" ); |
928 | Value *VOff = std::get<1>(in: getPtrParts(V)); |
929 | NewOff->addIncoming(V: VOff, BB); |
930 | } |
931 | copyMetadata(Dest: NewOff, Src: PHI); |
932 | |
933 | // Note: We don't eraseFromParent() the temporaries because we don't want |
934 | // to put the corrections maps in an inconstent state. That'll be handed |
935 | // during the rest of the killing. Also, `ValueToValueMapTy` guarantees |
936 | // that references in that map will be updated as well. |
937 | ConditionalTemps.push_back(Elt: cast<Instruction>(Val: Rsrc)); |
938 | ConditionalTemps.push_back(Elt: cast<Instruction>(Val: Off)); |
939 | Rsrc->replaceAllUsesWith(V: NewRsrc); |
940 | Off->replaceAllUsesWith(V: NewOff); |
941 | |
942 | // Save on recomputing the cycle traversals in known-root cases. |
943 | if (MaybeRsrc) |
944 | for (Value *V : Seen) |
945 | FoundRsrcs[cast<Instruction>(Val: V)] = NewRsrc; |
946 | } else if (isa<SelectInst>(Val: I)) { |
947 | if (MaybeRsrc) { |
948 | ConditionalTemps.push_back(Elt: cast<Instruction>(Val: Rsrc)); |
949 | Rsrc->replaceAllUsesWith(V: *MaybeRsrc); |
950 | for (Value *V : Seen) |
951 | FoundRsrcs[cast<Instruction>(Val: V)] = *MaybeRsrc; |
952 | } |
953 | } else { |
954 | llvm_unreachable("Only PHIs and selects go in the conditionals list" ); |
955 | } |
956 | } |
957 | } |
958 | |
959 | void SplitPtrStructs::killAndReplaceSplitInstructions( |
960 | SmallVectorImpl<Instruction *> &Origs) { |
961 | for (Instruction *I : ConditionalTemps) |
962 | I->eraseFromParent(); |
963 | |
964 | for (Instruction *I : Origs) { |
965 | if (!SplitUsers.contains(V: I)) |
966 | continue; |
967 | |
968 | SmallVector<DbgValueInst *> Dbgs; |
969 | findDbgValues(DbgValues&: Dbgs, V: I); |
970 | for (auto *Dbg : Dbgs) { |
971 | IRB.SetInsertPoint(Dbg); |
972 | auto &DL = I->getDataLayout(); |
973 | assert(isSplitFatPtr(I->getType()) && |
974 | "We should've RAUW'd away loads, stores, etc. at this point" ); |
975 | auto *OffDbg = cast<DbgValueInst>(Val: Dbg->clone()); |
976 | copyMetadata(Dest: OffDbg, Src: Dbg); |
977 | auto [Rsrc, Off] = getPtrParts(V: I); |
978 | |
979 | int64_t RsrcSz = DL.getTypeSizeInBits(Ty: Rsrc->getType()); |
980 | int64_t OffSz = DL.getTypeSizeInBits(Ty: Off->getType()); |
981 | |
982 | std::optional<DIExpression *> RsrcExpr = |
983 | DIExpression::createFragmentExpression(Expr: Dbg->getExpression(), OffsetInBits: 0, |
984 | SizeInBits: RsrcSz); |
985 | std::optional<DIExpression *> OffExpr = |
986 | DIExpression::createFragmentExpression(Expr: Dbg->getExpression(), OffsetInBits: RsrcSz, |
987 | SizeInBits: OffSz); |
988 | if (OffExpr) { |
989 | OffDbg->setExpression(*OffExpr); |
990 | OffDbg->replaceVariableLocationOp(OldValue: I, NewValue: Off); |
991 | IRB.Insert(I: OffDbg); |
992 | } else { |
993 | OffDbg->deleteValue(); |
994 | } |
995 | if (RsrcExpr) { |
996 | Dbg->setExpression(*RsrcExpr); |
997 | Dbg->replaceVariableLocationOp(OldValue: I, NewValue: Rsrc); |
998 | } else { |
999 | Dbg->replaceVariableLocationOp(OldValue: I, NewValue: UndefValue::get(T: I->getType())); |
1000 | } |
1001 | } |
1002 | |
1003 | Value *Poison = PoisonValue::get(T: I->getType()); |
1004 | I->replaceUsesWithIf(New: Poison, ShouldReplace: [&](const Use &U) -> bool { |
1005 | if (const auto *UI = dyn_cast<Instruction>(Val: U.getUser())) |
1006 | return SplitUsers.contains(V: UI); |
1007 | return false; |
1008 | }); |
1009 | |
1010 | if (I->use_empty()) { |
1011 | I->eraseFromParent(); |
1012 | continue; |
1013 | } |
1014 | IRB.SetInsertPoint(*I->getInsertionPointAfterDef()); |
1015 | IRB.SetCurrentDebugLocation(I->getDebugLoc()); |
1016 | auto [Rsrc, Off] = getPtrParts(V: I); |
1017 | Value *Struct = PoisonValue::get(T: I->getType()); |
1018 | Struct = IRB.CreateInsertValue(Agg: Struct, Val: Rsrc, Idxs: 0); |
1019 | Struct = IRB.CreateInsertValue(Agg: Struct, Val: Off, Idxs: 1); |
1020 | copyMetadata(Dest: Struct, Src: I); |
1021 | Struct->takeName(V: I); |
1022 | I->replaceAllUsesWith(V: Struct); |
1023 | I->eraseFromParent(); |
1024 | } |
1025 | } |
1026 | |
1027 | void SplitPtrStructs::setAlign(CallInst *Intr, Align A, unsigned RsrcArgIdx) { |
1028 | LLVMContext &Ctx = Intr->getContext(); |
1029 | Intr->addParamAttr(ArgNo: RsrcArgIdx, Attr: Attribute::getWithAlignment(Context&: Ctx, Alignment: A)); |
1030 | } |
1031 | |
1032 | void SplitPtrStructs::insertPreMemOpFence(AtomicOrdering Order, |
1033 | SyncScope::ID SSID) { |
1034 | switch (Order) { |
1035 | case AtomicOrdering::Release: |
1036 | case AtomicOrdering::AcquireRelease: |
1037 | case AtomicOrdering::SequentiallyConsistent: |
1038 | IRB.CreateFence(Ordering: AtomicOrdering::Release, SSID); |
1039 | break; |
1040 | default: |
1041 | break; |
1042 | } |
1043 | } |
1044 | |
1045 | void SplitPtrStructs::insertPostMemOpFence(AtomicOrdering Order, |
1046 | SyncScope::ID SSID) { |
1047 | switch (Order) { |
1048 | case AtomicOrdering::Acquire: |
1049 | case AtomicOrdering::AcquireRelease: |
1050 | case AtomicOrdering::SequentiallyConsistent: |
1051 | IRB.CreateFence(Ordering: AtomicOrdering::Acquire, SSID); |
1052 | break; |
1053 | default: |
1054 | break; |
1055 | } |
1056 | } |
1057 | |
1058 | Value *SplitPtrStructs::handleMemoryInst(Instruction *I, Value *Arg, Value *Ptr, |
1059 | Type *Ty, Align Alignment, |
1060 | AtomicOrdering Order, bool IsVolatile, |
1061 | SyncScope::ID SSID) { |
1062 | IRB.SetInsertPoint(I); |
1063 | |
1064 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1065 | SmallVector<Value *, 5> Args; |
1066 | if (Arg) |
1067 | Args.push_back(Elt: Arg); |
1068 | Args.push_back(Elt: Rsrc); |
1069 | Args.push_back(Elt: Off); |
1070 | insertPreMemOpFence(Order, SSID); |
1071 | // soffset is always 0 for these cases, where we always want any offset to be |
1072 | // part of bounds checking and we don't know which parts of the GEPs is |
1073 | // uniform. |
1074 | Args.push_back(Elt: IRB.getInt32(C: 0)); |
1075 | |
1076 | uint32_t Aux = 0; |
1077 | bool IsInvariant = |
1078 | (isa<LoadInst>(Val: I) && I->getMetadata(KindID: LLVMContext::MD_invariant_load)); |
1079 | bool IsNonTemporal = I->getMetadata(KindID: LLVMContext::MD_nontemporal); |
1080 | // Atomic loads and stores need glc, atomic read-modify-write doesn't. |
1081 | bool IsOneWayAtomic = |
1082 | !isa<AtomicRMWInst>(Val: I) && Order != AtomicOrdering::NotAtomic; |
1083 | if (IsOneWayAtomic) |
1084 | Aux |= AMDGPU::CPol::GLC; |
1085 | if (IsNonTemporal && !IsInvariant) |
1086 | Aux |= AMDGPU::CPol::SLC; |
1087 | if (isa<LoadInst>(Val: I) && ST->getGeneration() == AMDGPUSubtarget::GFX10) |
1088 | Aux |= (Aux & AMDGPU::CPol::GLC ? AMDGPU::CPol::DLC : 0); |
1089 | if (IsVolatile) |
1090 | Aux |= AMDGPU::CPol::VOLATILE; |
1091 | Args.push_back(Elt: IRB.getInt32(C: Aux)); |
1092 | |
1093 | Intrinsic::ID IID = Intrinsic::not_intrinsic; |
1094 | if (isa<LoadInst>(Val: I)) |
1095 | IID = Order == AtomicOrdering::NotAtomic |
1096 | ? Intrinsic::amdgcn_raw_ptr_buffer_load |
1097 | : Intrinsic::amdgcn_raw_ptr_atomic_buffer_load; |
1098 | else if (isa<StoreInst>(Val: I)) |
1099 | IID = Intrinsic::amdgcn_raw_ptr_buffer_store; |
1100 | else if (auto *RMW = dyn_cast<AtomicRMWInst>(Val: I)) { |
1101 | switch (RMW->getOperation()) { |
1102 | case AtomicRMWInst::Xchg: |
1103 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_swap; |
1104 | break; |
1105 | case AtomicRMWInst::Add: |
1106 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_add; |
1107 | break; |
1108 | case AtomicRMWInst::Sub: |
1109 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_sub; |
1110 | break; |
1111 | case AtomicRMWInst::And: |
1112 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_and; |
1113 | break; |
1114 | case AtomicRMWInst::Or: |
1115 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_or; |
1116 | break; |
1117 | case AtomicRMWInst::Xor: |
1118 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_xor; |
1119 | break; |
1120 | case AtomicRMWInst::Max: |
1121 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_smax; |
1122 | break; |
1123 | case AtomicRMWInst::Min: |
1124 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_smin; |
1125 | break; |
1126 | case AtomicRMWInst::UMax: |
1127 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_umax; |
1128 | break; |
1129 | case AtomicRMWInst::UMin: |
1130 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_umin; |
1131 | break; |
1132 | case AtomicRMWInst::FAdd: |
1133 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_fadd; |
1134 | break; |
1135 | case AtomicRMWInst::FMax: |
1136 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_fmax; |
1137 | break; |
1138 | case AtomicRMWInst::FMin: |
1139 | IID = Intrinsic::amdgcn_raw_ptr_buffer_atomic_fmin; |
1140 | break; |
1141 | case AtomicRMWInst::FSub: { |
1142 | report_fatal_error(reason: "atomic floating point subtraction not supported for " |
1143 | "buffer resources and should've been expanded away" ); |
1144 | break; |
1145 | } |
1146 | case AtomicRMWInst::Nand: |
1147 | report_fatal_error(reason: "atomic nand not supported for buffer resources and " |
1148 | "should've been expanded away" ); |
1149 | break; |
1150 | case AtomicRMWInst::UIncWrap: |
1151 | case AtomicRMWInst::UDecWrap: |
1152 | report_fatal_error(reason: "wrapping increment/decrement not supported for " |
1153 | "buffer resources and should've ben expanded away" ); |
1154 | break; |
1155 | case AtomicRMWInst::BAD_BINOP: |
1156 | llvm_unreachable("Not sure how we got a bad binop" ); |
1157 | } |
1158 | } |
1159 | |
1160 | auto *Call = IRB.CreateIntrinsic(ID: IID, Types: Ty, Args); |
1161 | copyMetadata(Dest: Call, Src: I); |
1162 | setAlign(Intr: Call, A: Alignment, RsrcArgIdx: Arg ? 1 : 0); |
1163 | Call->takeName(V: I); |
1164 | |
1165 | insertPostMemOpFence(Order, SSID); |
1166 | // The "no moving p7 directly" rewrites ensure that this load or store won't |
1167 | // itself need to be split into parts. |
1168 | SplitUsers.insert(V: I); |
1169 | I->replaceAllUsesWith(V: Call); |
1170 | return Call; |
1171 | } |
1172 | |
1173 | PtrParts SplitPtrStructs::visitInstruction(Instruction &I) { |
1174 | return {nullptr, nullptr}; |
1175 | } |
1176 | |
1177 | PtrParts SplitPtrStructs::visitLoadInst(LoadInst &LI) { |
1178 | if (!isSplitFatPtr(Ty: LI.getPointerOperandType())) |
1179 | return {nullptr, nullptr}; |
1180 | handleMemoryInst(I: &LI, Arg: nullptr, Ptr: LI.getPointerOperand(), Ty: LI.getType(), |
1181 | Alignment: LI.getAlign(), Order: LI.getOrdering(), IsVolatile: LI.isVolatile(), |
1182 | SSID: LI.getSyncScopeID()); |
1183 | return {nullptr, nullptr}; |
1184 | } |
1185 | |
1186 | PtrParts SplitPtrStructs::visitStoreInst(StoreInst &SI) { |
1187 | if (!isSplitFatPtr(Ty: SI.getPointerOperandType())) |
1188 | return {nullptr, nullptr}; |
1189 | Value *Arg = SI.getValueOperand(); |
1190 | handleMemoryInst(I: &SI, Arg, Ptr: SI.getPointerOperand(), Ty: Arg->getType(), |
1191 | Alignment: SI.getAlign(), Order: SI.getOrdering(), IsVolatile: SI.isVolatile(), |
1192 | SSID: SI.getSyncScopeID()); |
1193 | return {nullptr, nullptr}; |
1194 | } |
1195 | |
1196 | PtrParts SplitPtrStructs::visitAtomicRMWInst(AtomicRMWInst &AI) { |
1197 | if (!isSplitFatPtr(Ty: AI.getPointerOperand()->getType())) |
1198 | return {nullptr, nullptr}; |
1199 | Value *Arg = AI.getValOperand(); |
1200 | handleMemoryInst(I: &AI, Arg, Ptr: AI.getPointerOperand(), Ty: Arg->getType(), |
1201 | Alignment: AI.getAlign(), Order: AI.getOrdering(), IsVolatile: AI.isVolatile(), |
1202 | SSID: AI.getSyncScopeID()); |
1203 | return {nullptr, nullptr}; |
1204 | } |
1205 | |
1206 | // Unlike load, store, and RMW, cmpxchg needs special handling to account |
1207 | // for the boolean argument. |
1208 | PtrParts SplitPtrStructs::visitAtomicCmpXchgInst(AtomicCmpXchgInst &AI) { |
1209 | Value *Ptr = AI.getPointerOperand(); |
1210 | if (!isSplitFatPtr(Ty: Ptr->getType())) |
1211 | return {nullptr, nullptr}; |
1212 | IRB.SetInsertPoint(&AI); |
1213 | |
1214 | Type *Ty = AI.getNewValOperand()->getType(); |
1215 | AtomicOrdering Order = AI.getMergedOrdering(); |
1216 | SyncScope::ID SSID = AI.getSyncScopeID(); |
1217 | bool IsNonTemporal = AI.getMetadata(KindID: LLVMContext::MD_nontemporal); |
1218 | |
1219 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1220 | insertPreMemOpFence(Order, SSID); |
1221 | |
1222 | uint32_t Aux = 0; |
1223 | if (IsNonTemporal) |
1224 | Aux |= AMDGPU::CPol::SLC; |
1225 | if (AI.isVolatile()) |
1226 | Aux |= AMDGPU::CPol::VOLATILE; |
1227 | auto *Call = |
1228 | IRB.CreateIntrinsic(ID: Intrinsic::amdgcn_raw_ptr_buffer_atomic_cmpswap, Types: Ty, |
1229 | Args: {AI.getNewValOperand(), AI.getCompareOperand(), Rsrc, |
1230 | Off, IRB.getInt32(C: 0), IRB.getInt32(C: Aux)}); |
1231 | copyMetadata(Dest: Call, Src: &AI); |
1232 | setAlign(Intr: Call, A: AI.getAlign(), RsrcArgIdx: 2); |
1233 | Call->takeName(V: &AI); |
1234 | insertPostMemOpFence(Order, SSID); |
1235 | |
1236 | Value *Res = PoisonValue::get(T: AI.getType()); |
1237 | Res = IRB.CreateInsertValue(Agg: Res, Val: Call, Idxs: 0); |
1238 | if (!AI.isWeak()) { |
1239 | Value *Succeeded = IRB.CreateICmpEQ(LHS: Call, RHS: AI.getCompareOperand()); |
1240 | Res = IRB.CreateInsertValue(Agg: Res, Val: Succeeded, Idxs: 1); |
1241 | } |
1242 | SplitUsers.insert(V: &AI); |
1243 | AI.replaceAllUsesWith(V: Res); |
1244 | return {nullptr, nullptr}; |
1245 | } |
1246 | |
1247 | PtrParts SplitPtrStructs::visitGetElementPtrInst(GetElementPtrInst &GEP) { |
1248 | using namespace llvm::PatternMatch; |
1249 | Value *Ptr = GEP.getPointerOperand(); |
1250 | if (!isSplitFatPtr(Ty: Ptr->getType())) |
1251 | return {nullptr, nullptr}; |
1252 | IRB.SetInsertPoint(&GEP); |
1253 | |
1254 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1255 | const DataLayout &DL = GEP.getDataLayout(); |
1256 | bool InBounds = GEP.isInBounds(); |
1257 | |
1258 | // In order to call emitGEPOffset() and thus not have to reimplement it, |
1259 | // we need the GEP result to have ptr addrspace(7) type. |
1260 | Type *FatPtrTy = IRB.getPtrTy(AddrSpace: AMDGPUAS::BUFFER_FAT_POINTER); |
1261 | if (auto *VT = dyn_cast<VectorType>(Val: Off->getType())) |
1262 | FatPtrTy = VectorType::get(ElementType: FatPtrTy, EC: VT->getElementCount()); |
1263 | GEP.mutateType(Ty: FatPtrTy); |
1264 | Value *OffAccum = emitGEPOffset(Builder: &IRB, DL, GEP: &GEP); |
1265 | GEP.mutateType(Ty: Ptr->getType()); |
1266 | if (match(V: OffAccum, P: m_Zero())) { // Constant-zero offset |
1267 | SplitUsers.insert(V: &GEP); |
1268 | return {Rsrc, Off}; |
1269 | } |
1270 | |
1271 | bool HasNonNegativeOff = false; |
1272 | if (auto *CI = dyn_cast<ConstantInt>(Val: OffAccum)) { |
1273 | HasNonNegativeOff = !CI->isNegative(); |
1274 | } |
1275 | Value *NewOff; |
1276 | if (match(V: Off, P: m_Zero())) { |
1277 | NewOff = OffAccum; |
1278 | } else { |
1279 | NewOff = IRB.CreateAdd(LHS: Off, RHS: OffAccum, Name: "" , |
1280 | /*hasNUW=*/HasNUW: InBounds && HasNonNegativeOff, |
1281 | /*hasNSW=*/HasNSW: false); |
1282 | } |
1283 | copyMetadata(Dest: NewOff, Src: &GEP); |
1284 | NewOff->takeName(V: &GEP); |
1285 | SplitUsers.insert(V: &GEP); |
1286 | return {Rsrc, NewOff}; |
1287 | } |
1288 | |
1289 | PtrParts SplitPtrStructs::visitPtrToIntInst(PtrToIntInst &PI) { |
1290 | Value *Ptr = PI.getPointerOperand(); |
1291 | if (!isSplitFatPtr(Ty: Ptr->getType())) |
1292 | return {nullptr, nullptr}; |
1293 | IRB.SetInsertPoint(&PI); |
1294 | |
1295 | Type *ResTy = PI.getType(); |
1296 | unsigned Width = ResTy->getScalarSizeInBits(); |
1297 | |
1298 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1299 | const DataLayout &DL = PI.getDataLayout(); |
1300 | unsigned FatPtrWidth = DL.getPointerSizeInBits(AS: AMDGPUAS::BUFFER_FAT_POINTER); |
1301 | |
1302 | Value *Res; |
1303 | if (Width <= BufferOffsetWidth) { |
1304 | Res = IRB.CreateIntCast(V: Off, DestTy: ResTy, /*isSigned=*/false, |
1305 | Name: PI.getName() + ".off" ); |
1306 | } else { |
1307 | Value *RsrcInt = IRB.CreatePtrToInt(V: Rsrc, DestTy: ResTy, Name: PI.getName() + ".rsrc" ); |
1308 | Value *Shl = IRB.CreateShl( |
1309 | LHS: RsrcInt, |
1310 | RHS: ConstantExpr::getIntegerValue(Ty: ResTy, V: APInt(Width, BufferOffsetWidth)), |
1311 | Name: "" , HasNUW: Width >= FatPtrWidth, HasNSW: Width > FatPtrWidth); |
1312 | Value *OffCast = IRB.CreateIntCast(V: Off, DestTy: ResTy, /*isSigned=*/false, |
1313 | Name: PI.getName() + ".off" ); |
1314 | Res = IRB.CreateOr(LHS: Shl, RHS: OffCast); |
1315 | } |
1316 | |
1317 | copyMetadata(Dest: Res, Src: &PI); |
1318 | Res->takeName(V: &PI); |
1319 | SplitUsers.insert(V: &PI); |
1320 | PI.replaceAllUsesWith(V: Res); |
1321 | return {nullptr, nullptr}; |
1322 | } |
1323 | |
1324 | PtrParts SplitPtrStructs::visitIntToPtrInst(IntToPtrInst &IP) { |
1325 | if (!isSplitFatPtr(Ty: IP.getType())) |
1326 | return {nullptr, nullptr}; |
1327 | IRB.SetInsertPoint(&IP); |
1328 | const DataLayout &DL = IP.getDataLayout(); |
1329 | unsigned RsrcPtrWidth = DL.getPointerSizeInBits(AS: AMDGPUAS::BUFFER_RESOURCE); |
1330 | Value *Int = IP.getOperand(i_nocapture: 0); |
1331 | Type *IntTy = Int->getType(); |
1332 | Type *RsrcIntTy = IntTy->getWithNewBitWidth(NewBitWidth: RsrcPtrWidth); |
1333 | unsigned Width = IntTy->getScalarSizeInBits(); |
1334 | |
1335 | auto *RetTy = cast<StructType>(Val: IP.getType()); |
1336 | Type *RsrcTy = RetTy->getElementType(N: 0); |
1337 | Type *OffTy = RetTy->getElementType(N: 1); |
1338 | Value *RsrcPart = IRB.CreateLShr( |
1339 | LHS: Int, |
1340 | RHS: ConstantExpr::getIntegerValue(Ty: IntTy, V: APInt(Width, BufferOffsetWidth))); |
1341 | Value *RsrcInt = IRB.CreateIntCast(V: RsrcPart, DestTy: RsrcIntTy, /*isSigned=*/false); |
1342 | Value *Rsrc = IRB.CreateIntToPtr(V: RsrcInt, DestTy: RsrcTy, Name: IP.getName() + ".rsrc" ); |
1343 | Value *Off = |
1344 | IRB.CreateIntCast(V: Int, DestTy: OffTy, /*IsSigned=*/isSigned: false, Name: IP.getName() + ".off" ); |
1345 | |
1346 | copyMetadata(Dest: Rsrc, Src: &IP); |
1347 | SplitUsers.insert(V: &IP); |
1348 | return {Rsrc, Off}; |
1349 | } |
1350 | |
1351 | PtrParts SplitPtrStructs::visitAddrSpaceCastInst(AddrSpaceCastInst &I) { |
1352 | if (!isSplitFatPtr(Ty: I.getType())) |
1353 | return {nullptr, nullptr}; |
1354 | IRB.SetInsertPoint(&I); |
1355 | Value *In = I.getPointerOperand(); |
1356 | // No-op casts preserve parts |
1357 | if (In->getType() == I.getType()) { |
1358 | auto [Rsrc, Off] = getPtrParts(V: In); |
1359 | SplitUsers.insert(V: &I); |
1360 | return {Rsrc, Off}; |
1361 | } |
1362 | if (I.getSrcAddressSpace() != AMDGPUAS::BUFFER_RESOURCE) |
1363 | report_fatal_error(reason: "Only buffer resources (addrspace 8) can be cast to " |
1364 | "buffer fat pointers (addrspace 7)" ); |
1365 | Type *OffTy = cast<StructType>(Val: I.getType())->getElementType(N: 1); |
1366 | Value *ZeroOff = Constant::getNullValue(Ty: OffTy); |
1367 | SplitUsers.insert(V: &I); |
1368 | return {In, ZeroOff}; |
1369 | } |
1370 | |
1371 | PtrParts SplitPtrStructs::visitICmpInst(ICmpInst &Cmp) { |
1372 | Value *Lhs = Cmp.getOperand(i_nocapture: 0); |
1373 | if (!isSplitFatPtr(Ty: Lhs->getType())) |
1374 | return {nullptr, nullptr}; |
1375 | Value *Rhs = Cmp.getOperand(i_nocapture: 1); |
1376 | IRB.SetInsertPoint(&Cmp); |
1377 | ICmpInst::Predicate Pred = Cmp.getPredicate(); |
1378 | |
1379 | assert((Pred == ICmpInst::ICMP_EQ || Pred == ICmpInst::ICMP_NE) && |
1380 | "Pointer comparison is only equal or unequal" ); |
1381 | auto [LhsRsrc, LhsOff] = getPtrParts(V: Lhs); |
1382 | auto [RhsRsrc, RhsOff] = getPtrParts(V: Rhs); |
1383 | Value *RsrcCmp = |
1384 | IRB.CreateICmp(P: Pred, LHS: LhsRsrc, RHS: RhsRsrc, Name: Cmp.getName() + ".rsrc" ); |
1385 | copyMetadata(Dest: RsrcCmp, Src: &Cmp); |
1386 | Value *OffCmp = IRB.CreateICmp(P: Pred, LHS: LhsOff, RHS: RhsOff, Name: Cmp.getName() + ".off" ); |
1387 | copyMetadata(Dest: OffCmp, Src: &Cmp); |
1388 | |
1389 | Value *Res = nullptr; |
1390 | if (Pred == ICmpInst::ICMP_EQ) |
1391 | Res = IRB.CreateAnd(LHS: RsrcCmp, RHS: OffCmp); |
1392 | else if (Pred == ICmpInst::ICMP_NE) |
1393 | Res = IRB.CreateOr(LHS: RsrcCmp, RHS: OffCmp); |
1394 | copyMetadata(Dest: Res, Src: &Cmp); |
1395 | Res->takeName(V: &Cmp); |
1396 | SplitUsers.insert(V: &Cmp); |
1397 | Cmp.replaceAllUsesWith(V: Res); |
1398 | return {nullptr, nullptr}; |
1399 | } |
1400 | |
1401 | PtrParts SplitPtrStructs::visitFreezeInst(FreezeInst &I) { |
1402 | if (!isSplitFatPtr(Ty: I.getType())) |
1403 | return {nullptr, nullptr}; |
1404 | IRB.SetInsertPoint(&I); |
1405 | auto [Rsrc, Off] = getPtrParts(V: I.getOperand(i_nocapture: 0)); |
1406 | |
1407 | Value *RsrcRes = IRB.CreateFreeze(V: Rsrc, Name: I.getName() + ".rsrc" ); |
1408 | copyMetadata(Dest: RsrcRes, Src: &I); |
1409 | Value *OffRes = IRB.CreateFreeze(V: Off, Name: I.getName() + ".off" ); |
1410 | copyMetadata(Dest: OffRes, Src: &I); |
1411 | SplitUsers.insert(V: &I); |
1412 | return {RsrcRes, OffRes}; |
1413 | } |
1414 | |
1415 | PtrParts SplitPtrStructs::(ExtractElementInst &I) { |
1416 | if (!isSplitFatPtr(Ty: I.getType())) |
1417 | return {nullptr, nullptr}; |
1418 | IRB.SetInsertPoint(&I); |
1419 | Value *Vec = I.getVectorOperand(); |
1420 | Value *Idx = I.getIndexOperand(); |
1421 | auto [Rsrc, Off] = getPtrParts(V: Vec); |
1422 | |
1423 | Value *RsrcRes = IRB.CreateExtractElement(Vec: Rsrc, Idx, Name: I.getName() + ".rsrc" ); |
1424 | copyMetadata(Dest: RsrcRes, Src: &I); |
1425 | Value *OffRes = IRB.CreateExtractElement(Vec: Off, Idx, Name: I.getName() + ".off" ); |
1426 | copyMetadata(Dest: OffRes, Src: &I); |
1427 | SplitUsers.insert(V: &I); |
1428 | return {RsrcRes, OffRes}; |
1429 | } |
1430 | |
1431 | PtrParts SplitPtrStructs::visitInsertElementInst(InsertElementInst &I) { |
1432 | // The mutated instructions temporarily don't return vectors, and so |
1433 | // we need the generic getType() here to avoid crashes. |
1434 | if (!isSplitFatPtr(Ty: cast<Instruction>(Val&: I).getType())) |
1435 | return {nullptr, nullptr}; |
1436 | IRB.SetInsertPoint(&I); |
1437 | Value *Vec = I.getOperand(i_nocapture: 0); |
1438 | Value *Elem = I.getOperand(i_nocapture: 1); |
1439 | Value *Idx = I.getOperand(i_nocapture: 2); |
1440 | auto [VecRsrc, VecOff] = getPtrParts(V: Vec); |
1441 | auto [ElemRsrc, ElemOff] = getPtrParts(V: Elem); |
1442 | |
1443 | Value *RsrcRes = |
1444 | IRB.CreateInsertElement(Vec: VecRsrc, NewElt: ElemRsrc, Idx, Name: I.getName() + ".rsrc" ); |
1445 | copyMetadata(Dest: RsrcRes, Src: &I); |
1446 | Value *OffRes = |
1447 | IRB.CreateInsertElement(Vec: VecOff, NewElt: ElemOff, Idx, Name: I.getName() + ".off" ); |
1448 | copyMetadata(Dest: OffRes, Src: &I); |
1449 | SplitUsers.insert(V: &I); |
1450 | return {RsrcRes, OffRes}; |
1451 | } |
1452 | |
1453 | PtrParts SplitPtrStructs::visitShuffleVectorInst(ShuffleVectorInst &I) { |
1454 | // Cast is needed for the same reason as insertelement's. |
1455 | if (!isSplitFatPtr(Ty: cast<Instruction>(Val&: I).getType())) |
1456 | return {nullptr, nullptr}; |
1457 | IRB.SetInsertPoint(&I); |
1458 | |
1459 | Value *V1 = I.getOperand(i_nocapture: 0); |
1460 | Value *V2 = I.getOperand(i_nocapture: 1); |
1461 | ArrayRef<int> Mask = I.getShuffleMask(); |
1462 | auto [V1Rsrc, V1Off] = getPtrParts(V: V1); |
1463 | auto [V2Rsrc, V2Off] = getPtrParts(V: V2); |
1464 | |
1465 | Value *RsrcRes = |
1466 | IRB.CreateShuffleVector(V1: V1Rsrc, V2: V2Rsrc, Mask, Name: I.getName() + ".rsrc" ); |
1467 | copyMetadata(Dest: RsrcRes, Src: &I); |
1468 | Value *OffRes = |
1469 | IRB.CreateShuffleVector(V1: V1Off, V2: V2Off, Mask, Name: I.getName() + ".off" ); |
1470 | copyMetadata(Dest: OffRes, Src: &I); |
1471 | SplitUsers.insert(V: &I); |
1472 | return {RsrcRes, OffRes}; |
1473 | } |
1474 | |
1475 | PtrParts SplitPtrStructs::visitPHINode(PHINode &PHI) { |
1476 | if (!isSplitFatPtr(Ty: PHI.getType())) |
1477 | return {nullptr, nullptr}; |
1478 | IRB.SetInsertPoint(*PHI.getInsertionPointAfterDef()); |
1479 | // Phi nodes will be handled in post-processing after we've visited every |
1480 | // instruction. However, instead of just returning {nullptr, nullptr}, |
1481 | // we explicitly create the temporary extractvalue operations that are our |
1482 | // temporary results so that they end up at the beginning of the block with |
1483 | // the PHIs. |
1484 | Value *TmpRsrc = IRB.CreateExtractValue(Agg: &PHI, Idxs: 0, Name: PHI.getName() + ".rsrc" ); |
1485 | Value *TmpOff = IRB.CreateExtractValue(Agg: &PHI, Idxs: 1, Name: PHI.getName() + ".off" ); |
1486 | Conditionals.push_back(Elt: &PHI); |
1487 | SplitUsers.insert(V: &PHI); |
1488 | return {TmpRsrc, TmpOff}; |
1489 | } |
1490 | |
1491 | PtrParts SplitPtrStructs::visitSelectInst(SelectInst &SI) { |
1492 | if (!isSplitFatPtr(Ty: SI.getType())) |
1493 | return {nullptr, nullptr}; |
1494 | IRB.SetInsertPoint(&SI); |
1495 | |
1496 | Value *Cond = SI.getCondition(); |
1497 | Value *True = SI.getTrueValue(); |
1498 | Value *False = SI.getFalseValue(); |
1499 | auto [TrueRsrc, TrueOff] = getPtrParts(V: True); |
1500 | auto [FalseRsrc, FalseOff] = getPtrParts(V: False); |
1501 | |
1502 | Value *RsrcRes = |
1503 | IRB.CreateSelect(C: Cond, True: TrueRsrc, False: FalseRsrc, Name: SI.getName() + ".rsrc" , MDFrom: &SI); |
1504 | copyMetadata(Dest: RsrcRes, Src: &SI); |
1505 | Conditionals.push_back(Elt: &SI); |
1506 | Value *OffRes = |
1507 | IRB.CreateSelect(C: Cond, True: TrueOff, False: FalseOff, Name: SI.getName() + ".off" , MDFrom: &SI); |
1508 | copyMetadata(Dest: OffRes, Src: &SI); |
1509 | SplitUsers.insert(V: &SI); |
1510 | return {RsrcRes, OffRes}; |
1511 | } |
1512 | |
1513 | /// Returns true if this intrinsic needs to be removed when it is |
1514 | /// applied to `ptr addrspace(7)` values. Calls to these intrinsics are |
1515 | /// rewritten into calls to versions of that intrinsic on the resource |
1516 | /// descriptor. |
1517 | static bool isRemovablePointerIntrinsic(Intrinsic::ID IID) { |
1518 | switch (IID) { |
1519 | default: |
1520 | return false; |
1521 | case Intrinsic::ptrmask: |
1522 | case Intrinsic::invariant_start: |
1523 | case Intrinsic::invariant_end: |
1524 | case Intrinsic::launder_invariant_group: |
1525 | case Intrinsic::strip_invariant_group: |
1526 | return true; |
1527 | } |
1528 | } |
1529 | |
1530 | PtrParts SplitPtrStructs::visitIntrinsicInst(IntrinsicInst &I) { |
1531 | Intrinsic::ID IID = I.getIntrinsicID(); |
1532 | switch (IID) { |
1533 | default: |
1534 | break; |
1535 | case Intrinsic::ptrmask: { |
1536 | Value *Ptr = I.getArgOperand(i: 0); |
1537 | if (!isSplitFatPtr(Ty: Ptr->getType())) |
1538 | return {nullptr, nullptr}; |
1539 | Value *Mask = I.getArgOperand(i: 1); |
1540 | IRB.SetInsertPoint(&I); |
1541 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1542 | if (Mask->getType() != Off->getType()) |
1543 | report_fatal_error(reason: "offset width is not equal to index width of fat " |
1544 | "pointer (data layout not set up correctly?)" ); |
1545 | Value *OffRes = IRB.CreateAnd(LHS: Off, RHS: Mask, Name: I.getName() + ".off" ); |
1546 | copyMetadata(Dest: OffRes, Src: &I); |
1547 | SplitUsers.insert(V: &I); |
1548 | return {Rsrc, OffRes}; |
1549 | } |
1550 | // Pointer annotation intrinsics that, given their object-wide nature |
1551 | // operate on the resource part. |
1552 | case Intrinsic::invariant_start: { |
1553 | Value *Ptr = I.getArgOperand(i: 1); |
1554 | if (!isSplitFatPtr(Ty: Ptr->getType())) |
1555 | return {nullptr, nullptr}; |
1556 | IRB.SetInsertPoint(&I); |
1557 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1558 | Type *NewTy = PointerType::get(C&: I.getContext(), AddressSpace: AMDGPUAS::BUFFER_RESOURCE); |
1559 | auto *NewRsrc = IRB.CreateIntrinsic(ID: IID, Types: {NewTy}, Args: {I.getOperand(i_nocapture: 0), Rsrc}); |
1560 | copyMetadata(Dest: NewRsrc, Src: &I); |
1561 | NewRsrc->takeName(V: &I); |
1562 | SplitUsers.insert(V: &I); |
1563 | I.replaceAllUsesWith(V: NewRsrc); |
1564 | return {nullptr, nullptr}; |
1565 | } |
1566 | case Intrinsic::invariant_end: { |
1567 | Value *RealPtr = I.getArgOperand(i: 2); |
1568 | if (!isSplitFatPtr(Ty: RealPtr->getType())) |
1569 | return {nullptr, nullptr}; |
1570 | IRB.SetInsertPoint(&I); |
1571 | Value *RealRsrc = getPtrParts(V: RealPtr).first; |
1572 | Value *InvPtr = I.getArgOperand(i: 0); |
1573 | Value *Size = I.getArgOperand(i: 1); |
1574 | Value *NewRsrc = IRB.CreateIntrinsic(ID: IID, Types: {RealRsrc->getType()}, |
1575 | Args: {InvPtr, Size, RealRsrc}); |
1576 | copyMetadata(Dest: NewRsrc, Src: &I); |
1577 | NewRsrc->takeName(V: &I); |
1578 | SplitUsers.insert(V: &I); |
1579 | I.replaceAllUsesWith(V: NewRsrc); |
1580 | return {nullptr, nullptr}; |
1581 | } |
1582 | case Intrinsic::launder_invariant_group: |
1583 | case Intrinsic::strip_invariant_group: { |
1584 | Value *Ptr = I.getArgOperand(i: 0); |
1585 | if (!isSplitFatPtr(Ty: Ptr->getType())) |
1586 | return {nullptr, nullptr}; |
1587 | IRB.SetInsertPoint(&I); |
1588 | auto [Rsrc, Off] = getPtrParts(V: Ptr); |
1589 | Value *NewRsrc = IRB.CreateIntrinsic(ID: IID, Types: {Rsrc->getType()}, Args: {Rsrc}); |
1590 | copyMetadata(Dest: NewRsrc, Src: &I); |
1591 | NewRsrc->takeName(V: &I); |
1592 | SplitUsers.insert(V: &I); |
1593 | return {NewRsrc, Off}; |
1594 | } |
1595 | } |
1596 | return {nullptr, nullptr}; |
1597 | } |
1598 | |
1599 | void SplitPtrStructs::processFunction(Function &F) { |
1600 | ST = &TM->getSubtarget<GCNSubtarget>(F); |
1601 | SmallVector<Instruction *, 0> Originals; |
1602 | LLVM_DEBUG(dbgs() << "Splitting pointer structs in function: " << F.getName() |
1603 | << "\n" ); |
1604 | for (Instruction &I : instructions(F)) |
1605 | Originals.push_back(Elt: &I); |
1606 | for (Instruction *I : Originals) { |
1607 | auto [Rsrc, Off] = visit(I); |
1608 | assert(((Rsrc && Off) || (!Rsrc && !Off)) && |
1609 | "Can't have a resource but no offset" ); |
1610 | if (Rsrc) |
1611 | RsrcParts[I] = Rsrc; |
1612 | if (Off) |
1613 | OffParts[I] = Off; |
1614 | } |
1615 | processConditionals(); |
1616 | killAndReplaceSplitInstructions(Origs&: Originals); |
1617 | |
1618 | // Clean up after ourselves to save on memory. |
1619 | RsrcParts.clear(); |
1620 | OffParts.clear(); |
1621 | SplitUsers.clear(); |
1622 | Conditionals.clear(); |
1623 | ConditionalTemps.clear(); |
1624 | } |
1625 | |
1626 | namespace { |
1627 | class AMDGPULowerBufferFatPointers : public ModulePass { |
1628 | public: |
1629 | static char ID; |
1630 | |
1631 | AMDGPULowerBufferFatPointers() : ModulePass(ID) { |
1632 | initializeAMDGPULowerBufferFatPointersPass( |
1633 | *PassRegistry::getPassRegistry()); |
1634 | } |
1635 | |
1636 | bool run(Module &M, const TargetMachine &TM); |
1637 | bool runOnModule(Module &M) override; |
1638 | |
1639 | void getAnalysisUsage(AnalysisUsage &AU) const override; |
1640 | }; |
1641 | } // namespace |
1642 | |
1643 | /// Returns true if there are values that have a buffer fat pointer in them, |
1644 | /// which means we'll need to perform rewrites on this function. As a side |
1645 | /// effect, this will populate the type remapping cache. |
1646 | static bool containsBufferFatPointers(const Function &F, |
1647 | BufferFatPtrToStructTypeMap *TypeMap) { |
1648 | bool HasFatPointers = false; |
1649 | for (const BasicBlock &BB : F) |
1650 | for (const Instruction &I : BB) |
1651 | HasFatPointers |= (I.getType() != TypeMap->remapType(SrcTy: I.getType())); |
1652 | return HasFatPointers; |
1653 | } |
1654 | |
1655 | static bool hasFatPointerInterface(const Function &F, |
1656 | BufferFatPtrToStructTypeMap *TypeMap) { |
1657 | Type *Ty = F.getFunctionType(); |
1658 | return Ty != TypeMap->remapType(SrcTy: Ty); |
1659 | } |
1660 | |
1661 | /// Move the body of `OldF` into a new function, returning it. |
1662 | static Function *moveFunctionAdaptingType(Function *OldF, FunctionType *NewTy, |
1663 | ValueToValueMapTy &CloneMap) { |
1664 | bool IsIntrinsic = OldF->isIntrinsic(); |
1665 | Function *NewF = |
1666 | Function::Create(Ty: NewTy, Linkage: OldF->getLinkage(), AddrSpace: OldF->getAddressSpace()); |
1667 | NewF->IsNewDbgInfoFormat = OldF->IsNewDbgInfoFormat; |
1668 | NewF->copyAttributesFrom(Src: OldF); |
1669 | NewF->copyMetadata(Src: OldF, Offset: 0); |
1670 | NewF->takeName(V: OldF); |
1671 | NewF->updateAfterNameChange(); |
1672 | NewF->setDLLStorageClass(OldF->getDLLStorageClass()); |
1673 | OldF->getParent()->getFunctionList().insertAfter(where: OldF->getIterator(), New: NewF); |
1674 | |
1675 | while (!OldF->empty()) { |
1676 | BasicBlock *BB = &OldF->front(); |
1677 | BB->removeFromParent(); |
1678 | BB->insertInto(Parent: NewF); |
1679 | CloneMap[BB] = BB; |
1680 | for (Instruction &I : *BB) { |
1681 | CloneMap[&I] = &I; |
1682 | } |
1683 | } |
1684 | |
1685 | AttributeMask PtrOnlyAttrs; |
1686 | for (auto K : |
1687 | {Attribute::Dereferenceable, Attribute::DereferenceableOrNull, |
1688 | Attribute::NoAlias, Attribute::NoCapture, Attribute::NoFree, |
1689 | Attribute::NonNull, Attribute::NullPointerIsValid, Attribute::ReadNone, |
1690 | Attribute::ReadOnly, Attribute::WriteOnly}) { |
1691 | PtrOnlyAttrs.addAttribute(Val: K); |
1692 | } |
1693 | SmallVector<AttributeSet> ArgAttrs; |
1694 | AttributeList OldAttrs = OldF->getAttributes(); |
1695 | |
1696 | for (auto [I, OldArg, NewArg] : enumerate(First: OldF->args(), Rest: NewF->args())) { |
1697 | CloneMap[&NewArg] = &OldArg; |
1698 | NewArg.takeName(V: &OldArg); |
1699 | Type *OldArgTy = OldArg.getType(), *NewArgTy = NewArg.getType(); |
1700 | // Temporarily mutate type of `NewArg` to allow RAUW to work. |
1701 | NewArg.mutateType(Ty: OldArgTy); |
1702 | OldArg.replaceAllUsesWith(V: &NewArg); |
1703 | NewArg.mutateType(Ty: NewArgTy); |
1704 | |
1705 | AttributeSet ArgAttr = OldAttrs.getParamAttrs(ArgNo: I); |
1706 | // Intrinsics get their attributes fixed later. |
1707 | if (OldArgTy != NewArgTy && !IsIntrinsic) |
1708 | ArgAttr = ArgAttr.removeAttributes(C&: NewF->getContext(), AttrsToRemove: PtrOnlyAttrs); |
1709 | ArgAttrs.push_back(Elt: ArgAttr); |
1710 | } |
1711 | AttributeSet RetAttrs = OldAttrs.getRetAttrs(); |
1712 | if (OldF->getReturnType() != NewF->getReturnType() && !IsIntrinsic) |
1713 | RetAttrs = RetAttrs.removeAttributes(C&: NewF->getContext(), AttrsToRemove: PtrOnlyAttrs); |
1714 | NewF->setAttributes(AttributeList::get( |
1715 | C&: NewF->getContext(), FnAttrs: OldAttrs.getFnAttrs(), RetAttrs, ArgAttrs)); |
1716 | return NewF; |
1717 | } |
1718 | |
1719 | static void makeCloneInPraceMap(Function *F, ValueToValueMapTy &CloneMap) { |
1720 | for (Argument &A : F->args()) |
1721 | CloneMap[&A] = &A; |
1722 | for (BasicBlock &BB : *F) { |
1723 | CloneMap[&BB] = &BB; |
1724 | for (Instruction &I : BB) |
1725 | CloneMap[&I] = &I; |
1726 | } |
1727 | } |
1728 | |
1729 | bool AMDGPULowerBufferFatPointers::run(Module &M, const TargetMachine &TM) { |
1730 | bool Changed = false; |
1731 | const DataLayout &DL = M.getDataLayout(); |
1732 | // Record the functions which need to be remapped. |
1733 | // The second element of the pair indicates whether the function has to have |
1734 | // its arguments or return types adjusted. |
1735 | SmallVector<std::pair<Function *, bool>> NeedsRemap; |
1736 | |
1737 | BufferFatPtrToStructTypeMap StructTM(DL); |
1738 | BufferFatPtrToIntTypeMap IntTM(DL); |
1739 | for (const GlobalVariable &GV : M.globals()) { |
1740 | if (GV.getAddressSpace() == AMDGPUAS::BUFFER_FAT_POINTER) |
1741 | report_fatal_error(reason: "Global variables with a buffer fat pointer address " |
1742 | "space (7) are not supported" ); |
1743 | Type *VT = GV.getValueType(); |
1744 | if (VT != StructTM.remapType(SrcTy: VT)) |
1745 | report_fatal_error(reason: "Global variables that contain buffer fat pointers " |
1746 | "(address space 7 pointers) are unsupported. Use " |
1747 | "buffer resource pointers (address space 8) instead." ); |
1748 | } |
1749 | |
1750 | { |
1751 | // Collect all constant exprs and aggregates referenced by any function. |
1752 | SmallVector<Constant *, 8> Worklist; |
1753 | for (Function &F : M.functions()) |
1754 | for (Instruction &I : instructions(F)) |
1755 | for (Value *Op : I.operands()) |
1756 | if (isa<ConstantExpr>(Val: Op) || isa<ConstantAggregate>(Val: Op)) |
1757 | Worklist.push_back(Elt: cast<Constant>(Val: Op)); |
1758 | |
1759 | // Recursively look for any referenced buffer pointer constants. |
1760 | SmallPtrSet<Constant *, 8> Visited; |
1761 | SetVector<Constant *> BufferFatPtrConsts; |
1762 | while (!Worklist.empty()) { |
1763 | Constant *C = Worklist.pop_back_val(); |
1764 | if (!Visited.insert(Ptr: C).second) |
1765 | continue; |
1766 | if (isBufferFatPtrOrVector(Ty: C->getType())) |
1767 | BufferFatPtrConsts.insert(X: C); |
1768 | for (Value *Op : C->operands()) |
1769 | if (isa<ConstantExpr>(Val: Op) || isa<ConstantAggregate>(Val: Op)) |
1770 | Worklist.push_back(Elt: cast<Constant>(Val: Op)); |
1771 | } |
1772 | |
1773 | // Expand all constant expressions using fat buffer pointers to |
1774 | // instructions. |
1775 | Changed |= convertUsersOfConstantsToInstructions( |
1776 | Consts: BufferFatPtrConsts.getArrayRef(), /*RestrictToFunc=*/nullptr, |
1777 | /*RemoveDeadConstants=*/false, /*IncludeSelf=*/true); |
1778 | } |
1779 | |
1780 | StoreFatPtrsAsIntsVisitor MemOpsRewrite(&IntTM, M.getContext()); |
1781 | for (Function &F : M.functions()) { |
1782 | bool InterfaceChange = hasFatPointerInterface(F, TypeMap: &StructTM); |
1783 | bool BodyChanges = containsBufferFatPointers(F, TypeMap: &StructTM); |
1784 | Changed |= MemOpsRewrite.processFunction(F); |
1785 | if (InterfaceChange || BodyChanges) |
1786 | NeedsRemap.push_back(Elt: std::make_pair(x: &F, y&: InterfaceChange)); |
1787 | } |
1788 | if (NeedsRemap.empty()) |
1789 | return Changed; |
1790 | |
1791 | SmallVector<Function *> NeedsPostProcess; |
1792 | SmallVector<Function *> Intrinsics; |
1793 | // Keep one big map so as to memoize constants across functions. |
1794 | ValueToValueMapTy CloneMap; |
1795 | FatPtrConstMaterializer Materializer(&StructTM, CloneMap); |
1796 | |
1797 | ValueMapper LowerInFuncs(CloneMap, RF_None, &StructTM, &Materializer); |
1798 | for (auto [F, InterfaceChange] : NeedsRemap) { |
1799 | Function *NewF = F; |
1800 | if (InterfaceChange) |
1801 | NewF = moveFunctionAdaptingType( |
1802 | OldF: F, NewTy: cast<FunctionType>(Val: StructTM.remapType(SrcTy: F->getFunctionType())), |
1803 | CloneMap); |
1804 | else |
1805 | makeCloneInPraceMap(F, CloneMap); |
1806 | LowerInFuncs.remapFunction(F&: *NewF); |
1807 | if (NewF->isIntrinsic()) |
1808 | Intrinsics.push_back(Elt: NewF); |
1809 | else |
1810 | NeedsPostProcess.push_back(Elt: NewF); |
1811 | if (InterfaceChange) { |
1812 | F->replaceAllUsesWith(V: NewF); |
1813 | F->eraseFromParent(); |
1814 | } |
1815 | Changed = true; |
1816 | } |
1817 | StructTM.clear(); |
1818 | IntTM.clear(); |
1819 | CloneMap.clear(); |
1820 | |
1821 | SplitPtrStructs Splitter(M.getContext(), &TM); |
1822 | for (Function *F : NeedsPostProcess) |
1823 | Splitter.processFunction(F&: *F); |
1824 | for (Function *F : Intrinsics) { |
1825 | if (isRemovablePointerIntrinsic(IID: F->getIntrinsicID())) { |
1826 | F->eraseFromParent(); |
1827 | } else { |
1828 | std::optional<Function *> NewF = Intrinsic::remangleIntrinsicFunction(F); |
1829 | if (NewF) |
1830 | F->replaceAllUsesWith(V: *NewF); |
1831 | } |
1832 | } |
1833 | return Changed; |
1834 | } |
1835 | |
1836 | bool AMDGPULowerBufferFatPointers::runOnModule(Module &M) { |
1837 | TargetPassConfig &TPC = getAnalysis<TargetPassConfig>(); |
1838 | const TargetMachine &TM = TPC.getTM<TargetMachine>(); |
1839 | return run(M, TM); |
1840 | } |
1841 | |
1842 | char AMDGPULowerBufferFatPointers::ID = 0; |
1843 | |
1844 | char &llvm::AMDGPULowerBufferFatPointersID = AMDGPULowerBufferFatPointers::ID; |
1845 | |
1846 | void AMDGPULowerBufferFatPointers::getAnalysisUsage(AnalysisUsage &AU) const { |
1847 | AU.addRequired<TargetPassConfig>(); |
1848 | } |
1849 | |
1850 | #define PASS_DESC "Lower buffer fat pointer operations to buffer resources" |
1851 | INITIALIZE_PASS_BEGIN(AMDGPULowerBufferFatPointers, DEBUG_TYPE, PASS_DESC, |
1852 | false, false) |
1853 | INITIALIZE_PASS_DEPENDENCY(TargetPassConfig) |
1854 | INITIALIZE_PASS_END(AMDGPULowerBufferFatPointers, DEBUG_TYPE, PASS_DESC, false, |
1855 | false) |
1856 | #undef PASS_DESC |
1857 | |
1858 | ModulePass *llvm::createAMDGPULowerBufferFatPointersPass() { |
1859 | return new AMDGPULowerBufferFatPointers(); |
1860 | } |
1861 | |
1862 | PreservedAnalyses |
1863 | AMDGPULowerBufferFatPointersPass::run(Module &M, ModuleAnalysisManager &MA) { |
1864 | return AMDGPULowerBufferFatPointers().run(M, TM) ? PreservedAnalyses::none() |
1865 | : PreservedAnalyses::all(); |
1866 | } |
1867 | |